I Lost Faith in Kagi
d-shoot.net
d-shoot.net
- Search has been a breath of fresh air, I wish they dedicated more time to it.
- Orion...is ok? I use it off and on and it is fine but would rather have better search. The premise of the browser is nice but it feels like this could probably be a whole separate company or a purely open source endeavor. It has always been kind of clunky and not something I want to pay for.
- AI tools, I get the multiple pivots and I do believe that more recent advancements in ML/AI will make search a better experience but I do wish they had a little more focus.
- The t-shirts are weird and made me lose a lot of faith in their org. The way I look at it is instead of making their products better, they flushed a bunch of money down the drain for t-shirts and its offensive to paying customers who are paying because they enjoy the product and want it to become better, they don't want a t-shirt.
- I don't care about email, I don't care about other tools, make a great search experience first. Release all of the AI enhancements that you think will make sense, focus, focus, focus.
Edit: As I was adding my comment this post flagged and marked dead. Sometimes HN is weird.
1) I legit can't fathom going back to Google or any other search engine. I don't know what I'll do if they go under.
2) Investing in integrating AI into their search is absolutely vital and I like a lot of what they're doing there
3) Everything else, including the insanity of the t-shirts thing, is a complete waste of time and money. I don't understand what their strategy is if it isn't to set money on fire.
I've been mildly regretting not investing up until 5 minute ago when I read about spending 1/3 of that on the t-shirt factory.
> allocate[d] nearly a third of our investor-raised funds to produce and freely distribute 20,000 t-shirts
Though it sounds like they don't actually own "a t-shirt factory", but rather a t-shirt distributor.
The one thing I really hope they put more work into is searching for local news. That's one of the areas where I still have to turn to Google.
Presumably the tshirts are a marketing cost that they hope will lead to greater brand exposure and more subscribers.
What an own goal. I’m sure it made sense to them but I’m worried they don’t truly understand their customer base.
It seems like (again, t-shirts aside) Kagi is throwing a bunch of stuff at the wall to see what sticks. I hope they're having fun because I sure am.
What I want is a search engine that supports full-text queries with exact matches. This quite literally no longer exists anywhere, and maybe that's because it just doesn't scale. Nevertheless, I would find a lot more value in a search engine that returns exact matches. Someone will probably reply saying that Kagi, DDG, or The Google do exact matches with quotes, but this is not true. When it works, you've just gotten lucky. At best, it will filter out inexact matches, but that doesn't mean it will actually return every exact match in the index.
i’m glad they spent the time and effort on it.
Would you be upset if they had just donated that excess to charity?
Not upset in the slightest, I love Kagi search and want to see it continue. Merch is a solved problem and there was no reason to bring it in-house and make such a big announcement around it.
The difference they don't tell you about their internal accounting so you don't join the dots.
Start ups burn money on silly things like offices way too nice for what they need all the time. That's much closer to unethical than a company with no real duty to outsiders throwing away money.
Bradford Shellhammer (fab.com) wanting to speed run getting his United Global Services (invite only, elite tier, qualification criteria believed to be $50K annual base fare spend or 150K miles/year) so would fly back and forth between New York and Frankfurt every week, first class, while laying off employees left right and center.
But we are still allowed to criticize moves that we think are counter-productive or a waste.
Yes.
Do you think they charge too much?
Do they charge more than you think it's worth?
Or, at the bottom end, if their previous salaries were barely ramen-style living wages, I would be glad that they have become able to pay themselves enough to be comfortable.
As a paying customer, I want Kagi to succeed. I want Kagi's search offering to keep improving. Spending a couple hundred thousand of the company's money on t-shirts (one that I would receive, as I was a fairly early customer) sounds foolish to me, regardless of how much the founder is personally invested in the company, and regardless of whether or not he'll invest more of his own money to keep the company growing in the future, if needed.
I'm still bullish on Kagi's future, but things like this (and things mentioned in the linked article as a whole) make me a little worried.
> Would you be upset if they had just donated that excess to charity?
Probably! When I was at Twilio, we participated in GitHub's charity dodgeball tournament a few times (early last decade, I think). The cost of admission was $3000 per team, and would go to charity. After a couple years doing it, finance started getting uncomfortable with it. We were a private, unprofitable company (now Twilio is, of course, a public, unprofitable company), giving away money that our VCs had invested in us.
Initially I rolled my eyes, "just the bean-counters doing what they do best: whining about every bit of spend". But later, looking back, I realized they had a point. While $3000 wasn't a lot of money in the grand scheme of the company, what benefit was spending it actually providing the company? Ok, so 12 or so employees got to go and have a fun day at a rec center, boosting morale for them. We got our logo on the website for the tournament, which was maybe a little visibility/marketing. But was that really worth $3000 of our VC money? Maybe it was, but I don't think it's an obvious "yes".
What changes you have in mind to search functionality? I feel like the core search is rock solid as is, but they address search quality reports on their feedback forums all the time.
To me, the AI features (and specifically how they are only used when you opt in per query) are enhancing search, and the time they have been allocating to those features has continuously improved Kagis utility to me.
Note: I subscribe to Kagi Ultimate, so I use some AI features that are not available in the base plans.
It's just a matter of focus with a team of that size.
I tried Kagi and really enjoyed it but the pricing tier doesn't sit right with me, it's not that much better than DDG for my purposes. All these monthly subs start to add up. I'd be happier if there was a lifetime tier.
DDG and Kagi both use Bing data. That could be the reason
The modern, more polished Google search can make it literally impossible to find things that exist, because Google tries too hard to provide what it thinks you're looking for.
- Business listing search via maps is not a great experience. Maps and searching on maps are a more important endeavor than browser and email when thinking about the ecosystem.
- AI is definitely important but so far none of those features (afaik) have trickled down to non-ultimate users. From what I have seen, features have been removed from the regular plans.
- Remove reliance on using bing/google searches.
- Search is not a one and done operation.
Certainly not. I still get a decent amount of AI-generated blogspam in my results. Yes, it's great that Kagi offers me the option to manually block sites I don't want in results, but that's a workaround, not a solution, to the AI-generated spam problem.
I don't know if it's possible to detect this sort of crap automatically, but IMO this is the biggest threat to web search today.
Reverting the changes from around December that made it next to impossible to search for language-agnostic or English terms in another language.
Also reverting the changes over time that brought them closer to google or DDG and ignoring search terms unless you use verbatim or quote everything.
Kagi used to be about being explicit, but it’s slowly turning into the same "we know what you want to search for, so STFU" that all the other search engines are.
User since December 2021.
I have noticed the search has gotten worse in the ~7 months I've been using it, I started using Google more and more, and I was not planning to renew. I still use it, but after reading this article I'm definitely not going to use it again after my 1 year ends.
They aren't prioritizing the only feature I care about.
If they fail with all of the free marketing they’ve been gifted by this community I can only shake my head.
I'm a full Kagi shill. But I also want the stuff I like to remain stuff I like and reasonable criticism is the path there.
Kagi founder here and I want to clarify the train of thoughts around Kagi printing and giving away 20,000 t-shirts for its users.
- Kagi is not a typical VC funded startup.
- It is company I bootstrapped by going all in (meaning I put millions of dollars of my money into it).
- After all these years building it, we are lucky to have such incredibly passionate user community.
- That community is 100% responsible for Kagi's growth as a business through word of mouth (Kagi does no paid advertising).
- We are also famously taking a firm stance against ad-tech, so conventional advertising is not something I want to do.
- To do something as crazy as to start a company that builds a paid search engine and browser you obviously need to be thinking out of the box.
So combine all of this together and I thought that sending a t-shirt to all the people who supported us along the journey made a lot of sense.
The only thing I did not count on is how difficult will be to pull this off as I did not want to settle with less than premium quality for these t-shirts. As a result they will be delayed (my best guess is July/August) and I apologize for that to our users. In hindsight, we probably should have opted out for something easier to pull off (someone mentioned a billboard on 101, that would certainly be much easier).
This did not jeopardize Kagi's finances in any way at any point, nor I would do anything like that ever (as I said I am all in and have everything to lose, so I run a fiscally responsible business). In fact, Kagi has turned profitable recently.
This has also not impacted our ability to hire (we went from 10 people twelve months ago to 25+ now) and it did not impact our ability to ship a great product (check Kagi and Orion changelogs). I would venture to say that most Kagi users agree that Kagi is getting better and better every week with great speed.
So would I do it again? Well let's wait and see what we have in store for hitting 50,000 members mark :)
If the venture fails, you will ask yourself if you listened enough. Be proactive, address concerns, do not put yourself in a defensive position. Embrace change, be agile, and most importantly listen to your feedback.
Wish Kagi nothing but success, I would very much like a disruptor in this space. Best of luck to you and your team.
Or you know, that all cell phones had to have a physical keyboard. Until they suddenly didn't.
[Never tried Kagi, but let the man do his thing.]
I don't know if the OP got what they needed from this reply, but I assume I'm not alone in being impressed by the humility and candor of the response and developed much greater affinity for Kagi from some of the specifics of what were said.
I want more companies to have communicative, principled management that invites a sustainable base of like-minded customers/partners and fewer companies that pretend they can please 7B people by radically changing their product every 3 months.
I owned a business for 18 years. For 15 of those years it was my primary source of income. I valued feedback, tried my hardest to solicit as much of it as possible, and always took it to heart (though I had to always try and glean statistics from the sum of all feedback so that I was never spending resources on minority opinions).
What I read from the user was that the company created an optics problem. It wasn't whether the company was losing money or not, it was just that the user is choosing to support that company because they want a really good search engine, and the optics of divesting the company's resources into multiple projects makes it appear as if it could be the case that not enough focus is being spent on what really matters to that user.
What I read from the founder was that the optics issue went completely over his head and a complete dismissal of the user's concerns and feedback, along with a doubling down of the decisions made.
It's not a good look in my opinion. Even though the founder was polite and didn't say anything inappropriate, I would NEVER have responded to a customer of mine like that.
Hell, Orion is the first Webkit browser where FireFox and Chrome plug-ins work on iOS. If may seem like a misstep, but I see it as calculated. If Kagi search hopes to ever take on Google and Chrome, they need their own champion.
You are both right. Freediver laid out the vision, and some users are saying the vision isn't what the paying users are paying for. As someone who ran a business like this, GS is telling Freediver this should probably be something to give extra attention to and consciously decide is it the company the vision or the search product people are paying for?
The vision of a company shouldn't be much of a concern to paying customers of an existing product.
I think therein lies the problem for many people. If you're already at a stretch to justify paying for search, any deviation from your assumption of the correct decisions from the company will be magnified.
I think the most positive aspect of freediver's response is the implied dismissal of the above. - that their stubborness is genuine, not a more robotic, seemingly hollow, response of concern. As a marketing approach, I'm wondering if maybe that would give you less reach and more impact in general.
I'm not even arguing people should assume good faith until proven otherwise (I don't think we should, generally). Just that being so steadfast in one's assumption of bad faith is unwarranted.
If this were an occasional occurrence, I would agree with you, but it seems to happen systematically. Perhaps it's an exaggerated and counterproductive way of reacting.
People have the right to hold different opinions without feeling compelled to try to convince others of their own views or to be persuaded by others' opinions. Accepting that people can have differing viewpoints demonstrates respect for diversity of opinion and individual freedom of thought.
The concept of accepting different opinions without trying to convince each other implies respect for intellectual autonomy and freedom of thought for each individual, while still encouraging open dialogue and constructive engagement __when appropriate__.
As a woman who’s had bad experiences with men before, this kind of “no you owe me your time” behaviour usually triggers a “get away from this guy; fast” kind of response.
This doesn’t mean Kagi is necessarily bad or that this was what Vlad intended, but it does for sure reduce my faith in the project. Not enough to stop paying for Kagi, but I’ve definitely gone to the “raise shields, power phasers” stage.
If this is a regular "trigger point" in real life, I can understand that as a knee jerk, but there's no danger or need for immediate reaction via email, so why judge through that lens once you've identified you're doing so?
...and to be clear, Hacker News is not a representative sample of their customers.
I am a customer and I learned about Kagi here. I assume many people are on the same boat, so I wouldn't be so sure about that.
Perhaps I should have said "target customers" where I said "customers", I don't know. But it should not be surprising that "being an HN user" correlates strongly with "finds out about things on HN".
I would strongly bet their primary source of customers came via a HN referral.
We are literally their Target Customers.
If anything, I’m interested in what the evolution will look like as their customer base expands beyond HN types…
The only way a customer speaks is with money. If people like what you sell, you'll have more customers speaking with their wallet. If they don't then they tell you so by not purchasing what you sell. Internet commenters (such as myself) do not represent all customers or even a majority. People who are happy with a product usually see no reason to give feedback – especially when it's a small purchase. Likewise, people who hate your product wouldn't purchase it in the first place.
The cool part is that as people start leaving your product, complainers will become an even smaller minority, so you'll never have to second guess yourself. Maybe blame it on bullying?
For just about any business, if they were to ask their customers or the public at large what they want, the answer is usually "We want free stuff!". Cool to do if you're a politician, but bad business practice.
There's an old expression saying "the customer is always right", meaning that you can never blame the customers for how they spend or don't spend their money. If paying customers show a certain preference you better give it to them.
People who don't complain but silently drop are speaking with their wallets and that has to be listened to, as I said in my previous post. A business has to listen to customer spending behaviour and not listen too much to complainers. Normal people will give hotels awful reviews if it was raining on their vacation and great reviews if the weather was good and they had fun with their friends. Complaining is a past time to release some stress for many, and a pathological problem for a few. But when it comes to actually spending money is where the truth comes out.
Most people will not like your product and not buy your product, that's the large majority. That's why most normal businesses do not have the same reach as for example Apple or Toyota.
> The cool part is that as people start leaving your product, complainers will become an even smaller minority, so you'll never have to second guess yourself.
You can be sure that nobody second guesses themselves more than business leaders – especially if sales drop or stagnate. That doesn't mean that every complainer is right in their complaints.
As for Kagi there seems to be very many commenters online and in their feedback forums who believe that the main selling point of the service is privacy or extensive customisation. But I believe that the main selling point is search results quality and that everything else comes second. At least if they want to widen their customer base beyond computer hackers.
If you take a look at the Kagi feedback forums, there's almost every week somebody starting a thread where they demand that Kagi implements a very niche feature and then threatens to unsubscribe if they don't do it. Or demands a niche feature or they won't sign up. You can't listen too much to these people, you have to follow your own vision and if people agree with your decisions you'll see it in sales. If not, then you were wrong in your vision.
yeah well if I never pay money for kagi and never speak about anything how the hell is kagi supposed to know what they could do to get me to pay for them?
That doesn’t mean a business needs to acquiesce to every little demand. But “just listen to the money” is a horrible path for product improvement.
plenty of actionable complaints are just wild goose chases to nowhere.
I remember Daniel King's PowerPlayChess channel recently started promoting Kagi, doesn't this count as paid advertising or is this deal something else?
Delivering high-quality search over the entire Internet, higher quality than Google, is something so complex that even if you were literally the worlds smartest person and all the other Kagi employees were number 2 to number 26, there would still likely be stumbles at least once a year on something.
Because there’s like a million gotchas hidden along the path to just reliably matching Google search quality circa 2010 in the 2024 environment. Let alone delivering a high-quality browser, AI tools, etc., on top.
Google search has been bad for a long time. It's clear they serve their customers (advertisers) quite well, but as a user of their search, they're not particularly impressive.
The biggest problem is a problem of scale: being the biggest search provider means Google are targeted by SEO, so it's harder for Google to sift out the AI-generated garbage--Kagi just isn't involved in the arms race that Google is. But as a user that's not my problem; I'm not going to tolerate bad search results out of some sense of "fairness" to a corporation. And Kagi is delivering real user-centered features which are, frankly, obvious, i.e. Google should be embarrassed that they don't let you filter/prioritize domains or search within lenses like Kagi does.
But even the remainder will trip them up every now and then.
If these periods last only a few minutes it probably won't matter but if it lasts a few days or more then it's very likely to impact customer retention.
Did you not see the last part on your end?
Plus, if anything a small userbase makes it more difficult for quality search because the long tail is still effectively infinitely large, relative to the competencies of a single decision maker, but now there is only have one user searching for any random super niche topic maybe once a month, in total.
So they can't even a/b test or rely on customers reporting in on the real situation because it is too sparse.
In pursuing billions of global users across all demographics and trying to maximally monetize them through ads, Google is pursuing an entirely different measure of "search quality" than Kagi.
Google delivers their version of search quality when a rice farmer in Thailand and financier in the Bay Area both reach for Google when they want to find something online and then get distracted by an ad.
Meanwhile, Kagi gets their version right when they have a profitable base of happy customers. They can make different and more aggressive assumptions about the needs of their users, solicit and digest direct feedback about those assumptions, and optimize a product that delivers superb search quality for their niche.
They're completely different technical problems that only occasionally intersect. Their engineering teams aren't competing with each other.
So I don't see how Kagi can avoid having to deliver quality search results in millions of niches. Just at a very low frequency compared to Google.
Or are you suggesting to not bother with search quality past a certain lower threshold?
I don’t know about you, but the reason I’m paying for Kagi is because their search results are reliably better than Google.
That may be because all the spam optimizes for Google, but noneless they’ve already done the thing you say they should focus on.
As long as the search quality isn’t compromised it doesn’t matter to me what they do with my money.
Because of adverse selection?
Users that are users because of marketing are somehow different?
As does some sponsorship a la the VPN ads my kids see constantly in content heavy educational videos.
It is trying at all cost to make drivers lose focus of the road to see your advert, putting themselves and the people around them at risk of a road crash.
I love what you're doing and will continue to support you at your Professional tier as long as you continue doing what you're doing.
The number of times I've heard people extol the virtues of targeted ads on this site is absurd. I've even heard folks here say that Google ads are more helpful than the search results as if that's a good thing. And these are far more common comments here than comments in favor of actually returning good search results or aligning your income with user interests.
How often do you think the CEO pf Delta Airlines flies in first class versus on a private jet? My guess is only often enough to gin up a little PR.
And to be clear, I'm not even talking about being intentionally dishonest. AdTech workers likely believe the pro-ad propaganda they spout because they have to in order to live with themselves.
The thing you have to keep in mind is that HN is a very specific niche of the Internet. But for a slightly nerdy, not mass-market, product like mine (or Kagi) this niche is a great place to grow.
You just have to be mindful to see the feedback through the lens of the fact that you're talking to a niche audience and keep an eye on what a broader market might be looking for if that's where you're planning to go
You can find some people on HN who will say anything. That's the nature of a large distributed community.
But I also think the fact that there are people who would say that is a good thing. I despise ads, ad-tracking/privacy invasion, etc with a passion. I won't use anything that has ads. I won't even install most apps, even paid/premium, if they collect personal information. So you know my biases.
Their argument is a whole lot more nuanced (and better) than you are presenting it. The main point is that they find (or consider) ads to be useful for helping people find new products they might be interested in. And that does quite obviously happen sometimes (how often is very much up for debate though(. For them, the privacy tradeoff is worth it. I would guess that a portion of people who agree with that would fall into the "what do I have to hide if I'm not doing anythign wrong" camp. I think they're nuts, but it is an interesting viewpoint to consider that is a lot more nuanced and contextual then just "I enjoy targeted ads."
Honestly I get the T-shirt part this way. You got to Doo crazy stuff as a start-up. I also get that you try ai stuff. As long as you keep up de search.
However what scares me is the apparent lack of knowledge about privacy, gdpr and what is PII in a product that, to me, is all about privacy. Have one person in the company be an expert in privacy and GDPR etc and use their insights, since it is critical for your right of existence.
They'll likely discover that GDPR is not that optional as soon as a customer (or a competitor with a grudge!) reports them to their relevant national privacy/personal data protection authority, after which they'll get to have a very uncomfortable conversation where they will not be able to use those arguments
(Do you regularly check to make sure you're obeying laws in countries you don't ever intend to visit? No? Then why should Kagi?)
I suppose a serious breach of regulations, and if Kagi decided to ignore fines, apart from a bad reputation, could ultimately lead to things like judicial decisions of blocking access to the website or blocking payments for EU customers.
The scope of GDPR is clearly including businesses operating from the US, but has any company registered only in the US ever been fined by EU?
This means they're exposed to GDPR not only indirectly by serving EU citizens but also directly by operating within the EU.
AI tools that Kagi uses like vector database search and LLM connectors are already open source.
You can find these in our Github:
Quick (but difficult) question: do you foresee there arising a reasonably reliable way to filter out the coming wave of ai spam? I’m told that half of Twitter is bots talking to each other at this point, and I’m sure this is coming to other media as well. Eg, massive, massive waves of content marketing, sock puppets, etc.
Is there reason to be optimistic that you or other actors will be able to sift through it?
That's as good as 0% accuracy.
And AI text will keep changing.
I still don't agree with the shirts and I think the overarching point is the shirts seem like a common theme of trying to do too much. I hope my thinking is not true and I wish the best success because I love Kagi.
After all this very valid, very sensible feedback, you're commenting here saying "you need to be thinking out of the box" and trying to justify all the time, money and energy you spent on those t-shirts. Your customers are complaining because they want you to succeed. If they didn't care about you, they'd just cancel the subscription and move on. And your response to it is "nah, what we did was right" and not "yeah, maybe we shouldn't have spent all that resources on a stupid t-shirt that nobody wanted"?
I just don't get it. And what are you gonna do when you hit 50K members? Are you planning to send an entire wardrobe (from undies to a suit) to all your 50K customers (assuming IF you ever hit that many customers)?
Oh, and nitpick: It's outside the box, not out of the box...
The Founder of the company decided to pour in his own money worth of millions ( assuming that is true ) into the company. He bootstrapped it.
They were on track to becoming profitable. And now, as the replies shows they are now profitable.
He decided to spend some money to buy everyone a t-shirt as a gift.
And all I see is not thank you but anger and frustrations. Pointing to waste of time as if the product is going downhill or not iterating fast enough. But in fact according to changelog they are doing pretty well.
Spending my own money to say thank you in my own way and all I got was, all these negativity.
I just don't get it.
Read the article, the author clearly lays out how Kagi is scattering their attention, money and energy on things that do not matter one bit.
He decided to spend some money to buy everyone a t-shirt as a gift.
A gift that nobody asked for, a gift that has zero impact, a gift that took their attention, time and money away from their core product.
Let me repeat - people are complaining because they care, because they want Kagi to succeed. Google search has gotten so bad that people are desperate for an alternative. Instead of taking it as "people are mean to us", perhaps Kagi should take it as "let us listen to our customers, they really do want us to succeed". If you see it from the perspective of a Kagi fan/customer, your opinion might change
To move my mother to Kagi, I had to install Brave Browser on her android cell phone, make it the default browser on Android, change Brave's default search engine and create a desktop shortcut to it.
Android is ubiquitous in Brazil. I won't be able to move much people to Kagi with a process this involved.
Happy to chat further about this topic if you wish, privately or publicly.
I didn't flag it, but I came close just because the tone of the piece is so sensational and needlessly aggressive. I left it put because it's the first negative Kagi piece I've seen and I didn't want to silence an alternative perspective, but the quality was definitely below what I hope to see on the front page.
Things I love and can't live without:
- When I search for something, I don't have to deal with weeks of whatever I searched for coming up in ads on every web page I visit.
- I don't feel like "the man" is snooping on me in some sort of weird dark social credit score thing. (I literally got a call from Google once offering me a job based on what I'd been searching for. Flattering, but totally freaked me out)
- The quality is good for non-local things
- I'm the customer, not the product
- That makes things like blocking or enhancing sites possible
What I'd like to see improve:
- I don't want AI. I don't want summaries, I don't want hallucinations, I don't want assistants. I don't want it.
- Local results and map integration. When I click on a local result, actually having the map go to the result I clicked on. Currently this doesn't work well.
- Hours for local businesses.
I find myself still going to google for these things, and while it doesn't seem like a lot, aside from work stuff those kinds of searches are probably 80% of what I need. Where can we go for dinner tonight that's near by and still open? Who has all-you-can-eat deals near by? Where can I find some floating shelves to put in my office near by?
Those are all examples of things that Google does really well, and I don't have much luck with on Kagi.
I agree with the author that I'd rather see the quality there improve before AI features.
Quality is there for the most part, IME, but I definitely agree that their local features need a LOT of work.
Their other projects are not interesting or useful to me, but so far I can simply ignore them. Yes, on some level I wish they'd focus and quit wasting money and energy on things I don't care about, but that's really not my affair.
The one growing reservation I have is with regard to Vlad's/Kagi's actual, boots-on-the-ground approach to privacy. Kagi necessarily has the ability to know more about me than almost any other company. I want to see them demonstrate strong and unwavering commitment to respecting and protecting my privacy - through policy, technology, and careful and continuous vetting of partners. Expressed disinterest in collecting or capitalizing on my data is not enough, and seeing Vlad's communications in which he casually shrugs or responsibility-shifts to a third-party heightens my concern.
For now, I remain a customer - but a wary one. I've stopped actively recommending Kagi personally and professionally because as a privacy advocate, it increasingly feels irresponsible to do so.
The founder posted comments on hn assuring that they take privacy seriously, and I believe him, but most commercial companies (including the big ones like Microsoft) also claimed to take privacy seriously. Look at what they are doing right now: Blatant violations and even more blatant lies.
Search is a deeply personal activity. It can reveal far more information about the user than financial statements, health records, privileged attorney information, or library reading lists. Kagi therefore must _at least_ meet the same sanctity, privilege, and protection standards afforded to those parts of life. At present, Kagi does not meet these standards through technical means, and governing laws certainly fall short of compelling Kagi to meet either.
So while I appreciate what Kagi is trying to do and wish them success, I cannot see myself using it in its current form. Local (private) LLMs and fact checking through search engines that aren't tied to my PII simply provide a superior experience. At present, it's simply impossible for people like me who want better search and are willing to pay for it to become customers of Kagi. I find that to be a real shame :(
And to continue down the road of AI proficiency, Kagi will need to retain a lot of data.
Very annoying being a hobbyist website in the middle of it I'll tell you that much
I paid for a good search engine that respects my queries and does not try to outsmart me. The more Kagi focuses on AI, and making an """intelligent""" search engine, basically replicating Google's missteps, is the day I stop giving them money. I've already been noticing some of my keywords are being ignored or reinterpreted. Please stop that.
I don't care about email either. I am paying Fastmail for it, and I certainly know better than to attach my search history to my email account, especially when it's from an AI company. Is the goal here to copy Google?
To all startup owners: there is more to software quality and user experience than trying to fit the AI buzzword in anything you do. Stop following the hype and focus on building a damn good product.
My own experience has been that what I get month to month is worth what I pay. If the project is sustainable, then I'll get to enjoy it into the future. If not, I'll get to enjoy it while I can.
A search engine isn't like an email provider or even a web browser, there's basically no lock in that makes transitioning later difficult if something changes for the worse.
Sure, but what happens with your information after that is also very important. What's for me very concerning after reading the article is not a T-Shirt factory or burning budget, but the their attitude towards privacy.
I did trust that, before reading this article. The founder's attitude toward privacy -- based on what he's said in the Discord -- worries me, and some of that trust has been eroded.
Personally I think the t-shirts look great and I'd love one.
The only thing that concerned me in that article was the disregard and / or misunderstanding of GDPR.
This take feels more like being upset about one individual's (Vlad) personal opinions about privacy and politics. But in my opinion, it fails to realize that assigning one person's views to an entire organization is a fallacy. Even if they are the leader.
As a service, I like Kagi. Both in principle, and in practice. I find the "summarize this page" feature to be very useful. I also like the idea of paying for value, rather than being forced to feed the advertising beast. So I pay for value. If it stops being valuable, I will stop paying. I care about privacy, but I also realize that we live in a world where there are serious limits on the amount of privacy that can be expected. So I have to just do the best I can with what is available. Kagi is at least an improvement on the standard "eyeballs are the product" business model.
More generally though, word of mouth is a good place to start but it maxes out quickly, especially for niche products. There will need to be some support from other channels. Even just putting the name of the company on the t-shirt would have supported it a bit.
And Vlad didn't even say anything that crazy from a political perspective. "News should not only be about politics" is super reasonable, and I found myself agreeing with him much more than the person he was talking to.
While no one can truly be unbiased, I want them to try. Arguments against this are just sophist nonsense, and I'm not convinced they are made in good faith.
Someone who believes in objective truth believes their own viewpoint to be the only valid viewpoint, as indeed we see in the quotes in TFA.
So when some local newspaper reports about some random old person in an old-age home that's political?
Please, take this destructive attitude and reassess it.
Not all things are political and making them so actively makes the world worse.
Political isn't always negative but the news is always political. All news, including good sources like NPR, is promoting a point of view. Nothing is impartial even if that's the goal. Ignoring this doesn't change it.
t-shirts are something that people think they can understand, so they speak most at length about it compared to the other things Kagi is doing.
You're paying, what, 10, 25 USD - are you getting a good service for it? If not, unsubscribe, if yes, what's the problem? Sounds like they're profitable now, so little risk of the service dissapearing.
Unnecessary drama by people who live for drama. My only advice for Vlad would be to not get caught up in it.
Based on the exchanges, Vlad is both extremely combative and unwilling to accept the possibility that he is wrong (which he is here).
Being aggressively wrong is no way to go through life. Vlad should be more humble, and open to being wrong, rather than being unnecessarily belligerent.
Put more simply: it takes a weird, broken logic to find fault in the idea that a person who won't stop emailing you, after being told to, isn't "combative".
The email Lori sends explicitly asking him to stop emailing is then followed up with some last-wordism "for the record" nonsense. Only on the extremely online internet do people consider someone the aggrieved party after they write a screed against a product or business, then close the conversation with representatives of that business with essentially a don't @ me and some last-wordism. It's terrible journalistic practice. It's a net negative in social and community engagement. I don't see why doing it over online spaces gives the author a pass here.
If the founder wants to respond they can write a respectful blog post and put it on the fucking homepage. They don't have a right to harangue the author via email.
And why is that, exactly? They do have this “right” and they exercise it.
Or if that's the words that would've been chosen - I would agree to you.
But if you mix those words with extra message, then no. A reply to this new message is warranted.
E.g. if you add a reason and that reason is unreasonable - it's warranted to address that and reply to you. Either do a request without attached strings, or block. Don't write extra conditions/reasoning and then complain that someone doesn't agree with you on those and kept messaging.
Lori's emails are deliberately designed to goad Vlad into replying, just to act indignant when he does.
This is not goading, this is telling someone to fuck off into the sun. If he wants to respond he can respond on his product with a blog post. His audience already dwarfs hers anyways, he Streisand-effect'ed himself because he's clearly got some narcissism issues.
The exchanges all read like Vlad derives a lot of self-esteem from being right, which isn't as good as deriving it from ability to learn when wrong.
But, Vlad definitely should have stopped when Lori responded that they didn’t want to have a conversation at all. If for no other reason than they were a lost cause.
The customer isn't always right, but often is, like in this case. If you're a CEO, best to just pipe down, be humble, and listen to customers. Being open to being wrong is a nice plus, but either way, people will like you more if you appear to listen instead of argue. Even when you're right!
tl;dr: this isn't an internet argument between two otherwise-equal random strangers, this is a CEO talking down to a customer while being objectively wrong, which is 2x bad.
Instead of actually educating himself, he just argues that he's not wrong. I could easily see Lori being sick of the frustration of having to deal with that and just say "ok, nope, this conversation is done".
> They just put the company on blast and aren’t willing to even hear the other side of things.
I don't think that's what actually happened?
I think that if he had different type of personality then he wouldn't start this company - a regular, humble, humiliated, developer would just tremble, sweat and shiver at the thought of starting business straight competing with core Google, MS products. He needs to be believer and confident to pull this. Also almost all leaders of great and (now) big companies seem to be type of people that regular Joe not necessarily would enjoy to be friends with.
There is a great strength in "I might be wrong here, I'd like to learn more" than makes even the most hardheaded wrongness look weak, and if you look at the history of effective CEOs, you'll likely notice that inability to entertain the possibility you might be wrong tends to be a liability.
People discuss Apple's commitment to privacy and if it is real or adequate.
There are costs other than direct monetary. We're still "paying" for it, just via ads, sponsored results, etc.
people pay with wasted time and "cognitive load" because of the interstitial ads, and to decipher biases in presented data, though, too.
(i see a sibling comment is similar, but didn't mention wasted time, so leaving this here)
He reached out to her via email to set up a call. She demurred and asked him to stop contacting her, he persisted and wrote a petulant novella of an email. She asserted that he stop contacting her again. He seems to have finally taken the hint.
This is a guy who seems like he can't stand to be wrong about anything, not a business I would bet on with my wallet.
American culture and customer support likes to blow smoke up your ass while saying no. Other cultures and Vlad don’t do that.
I'm a huge fan of kagi and have been paying for it for as long as paying for it has been possible - that said, I think the author is spot on about the long term viability of the project considering their limited funding, limited employees, very wide (yet unproven) interests AND a leader who's maybe not so receptive to feedback.
For example I was part of the Orion beta and I left feedback in the discord that it took ~30 seconds on the then top of the line iPhone (13 Pro Max?) to load the interface which made it hard to use and I thought it was unreasonably slow and he said something like "that's not slow it's totally reasonable" and since then I decided it wasn't worth leaving any more feedback and have since left the community.
Otherwise, ' "reasonable default" for lots of text ' is something that browsers provide, using the "system" fonts. Applying a font-family to the entire html or body tags will do the job, because system fonts don't need to download or load into the browser. And since you can even specify the specific system font you want to use, you have a few options like serif or sans-serif.
All of that aside, if I applied a system font and your screen reader applies a different one, what was the point of the extra css? So that's my guess as to why people do this because, like you, I find it very hard to read.
If you're curious, though, Firefox has a built-in reader mode and I think Safari does, too. Last I checked, Chrome's was behind a flag. And then, of course, there are extensions (but extensions to read plain HTML docs seems exactly backwards, so...)
Seriously: no margins on the images and the images all different widths. No human being would lay out a mixed-media document like this on purpose if they expected other human beings to consume it easily.
(This reflects not so much on the author as on how fascinatingly bad the UX of unstyled HTML is. I remember when things looked like this and we were just used to it because there wasn't anything else on the web).
(which results being hard to copy from)
that Kagi founders use it is a huge red flag that their values and mine do not align.
I had to, just to get past the first couple sentences.
Sadly no reader view can divine where paragraphs should be, but aren't. This is just lazy editing.
The authors replies seem pretty rude (or at least somewhat aggressive / dismissive). Kagi is Vlads baby and I could imagine he would care and try to explain when he thinks someone has the wrong idea. However to the author - it’s just another service he doesn’t use anymore.
His business collects email addresses, which is a process. Under GDPR, this process must be documented, users must be given their data on request (even if it just contains an email address, but usually it also contains the signup date for example as a proof for their data processing consent) and users must be informed about their rights to correct or delete such data.
He comes off totally as the "trust me bro" guy with zero respect for a different perspective and doesn't seem to be interested in changing his (objectively wrong) opinion. It is almost laughable, because "is email PII" has been discussed a million times since the introduction of the GDPR that you must've lived under a rock to dismiss it like Vlad did.
Point is: if the business requests an email address, many people will provide their real email adress and your business needs to document this process under GDPR. I just checked. The signup form doesn't say "please give a FAKE email address", it just says "email address".
If a user provides a real email address, Kagi must respond to GDPR Art. 15 requests by providing...that same email adress. Might sound silly, but usually, there is other data associated with this. Usually, at least the timestamp of the signup. If a business is really GDPR compliant, it will offer a download option for stuff like user settings and so on.
Or, if the user signed up and later deleted the account, his email should explicitly NOT show up when asking for personal data.
See, it is about documenting the process, whether the outcome is "here is your email address you just asked for" or "we don't have any data on you". And Vlad says that this process is irrelevant for Kagi, while it is not.
I've made a bunch of SAR's, including pre-GDPR and I've never received one that contained my user settings, so that seems pretty normal.
The whole PII convo seems incredibly asinine though, "PII" is not a thing in the GDPR. Personal data is[1], but that's not the same thing.
If Kagi keep a record of searches performed by a user, that's something that a SAR should be used for, but the whole convo just misses the mark entirely.
[^1]: See article 4.1 https://gdpr-info.eu/art-4-gdpr/
That is odd because the user has right to get ALL the data they want that is stored about them in the service.
If he's positioning his company to challenge that law when he runs afoul of it, that's a choice they can make but it's a business risk (and IANAL, but... Probably one they'll lose).
> Personal emails are PII. But you can register to Kagi with a random email, and that is not PII.
Company has no means to identify what is the difference between personal or random email. Everything can be and must be treated equally from privacy perspective.
Vlad comes off as fairly unhinged here.
In my eyes this rationale would make sense if there was no backstory to this. If there was no preceeding blogpost, I'd consider Vlads messages pure spam.
But the context here is different: The author wrote a very critical, and clearly opinionated blogpost. There was clear intention in engaging with this subject.
Now the author seems to want to avoid responsibility, while Vlads attempt to react to a public hit piece with a respectful conversation was honestly the best way to handle this.
That's the fundamental premise of telling people that they are sealioning.
Not everyone agrees with it (I suspect age plays more a role than anything else).
Your historical example doesn't really map very well to today, because control over the ability to put some text somewhere that others can read it is very, very different than it was historically.
None of this excuses the Kagi CEO's failure to back off when asked/told to. They should just have used their own blog or equivalent to respond.
Still, generalizing to a broad claim about raising an issue in public creating no future obligations seems somewhat wrong to me. You don't have to speak in public about anything at all. For me, your choice to do so creates some limited obligations towards future engagement (though I'm not sure quite where the limits lie).
Yes, and then that engagement - which very much took place - did not give the author any confidence that FURTHER ENGAGEMENT (via email) would change the situation.
If I talk to you back and forth about an issue I have and feel like I'm talking to a brick wall, so I then write a critical review based on those issues, why should I be forced to not be a brick wall, in return? If Vlad wants someone to listen to him, he should probably take some time to engage with (not just 'listen to') what is being said on it's fundamental merits (not whatever surface level bit he wants to latch on to).
Recontextualizing an issue is not addressing it. Explaining an issue is not addressing it. Describing a paradigm that contextualizes an issue is not addressing it.
Vlad saw something critical of his hard work and wanted to put in the effort to clarify his stances and mend a relationship. I can absolutely understand that, your work is a reflection of yourself and nobody wants to be judged on misunderstanding. He might've even felt like he let someone who cared about Kagi down and wanted to make it right. Again, all understandable!
However, twice, the blog post author said they did not want to engage. At this point, regardless of how you feel about what was said, you should probably move on; they said their piece, you tried to engage, they rebuffed, oh well, do something else! To continue on is both incredibly annoying and a bit unhinged.
If Vlad absolutely felt like he needed to respond to this, he should've digested the main points of the original blogpost, reflected on them, and written his own blog post to a more general audience. Not necessarily in _response_ to the author, but understanding that more people probably feel this way as well and want to hear clear answers. Perfect examples of this would be an "Our stance on privacy" or "How we're ensuring Kagi's future," again factoring in the criticism from the author.
I write all of this as someone who pays for and likes Kagi. I think it's a good product, if a bit scattered at times. But the blog post does hit on some concerns that I have (privacy being the biggest) and seeing the follow up leaves a bad taste in my mouth.
But beyond that, there's some irony in that exchange. If Vlad had simply stopped engaging when Lori asked, it would indeed make Lori seem like more of an asshole for rejecting an appeal to have a simple conversation. But then Vlad transgressed that wish, making Lori's case about not wanting to engage.
If anything the replies in that Mastadon thread make the author and others appear petty, combative and immature imo, and I do not say that as someone who agrees with all Vlad's perspectives.
I'm all for generally leaving people alone, and being civil, but context please.
You don't get to open a salvo against someone, while pretending you're above interaction with them, then play the victim when they respond and universally and unilaterally dictate terms while always trying to get the last word in.
Anyway, I just think that people do things in online discussions that they wouldn't do to someone's face. And that tends to be a bad thing for reasonable discourse.
The Internet has turned us into sociopaths.
In the thread linked above, I think his reasoning for posting the email is reasonable. I find this similar to what the Apollo dev did when discussing with Reddit people. If he didn't record the conversation or make it public, his words could have been twisted.
Ummm dating and breakups?
Imagine doing this in the offline world. How well would this kind of behavior go over with people at the grocery store, do you think? Why is it acceptable online to behave like this?
The internet levels the playing field so the random person has the power to post criticism of the more powerful person and be heard. It doesn’t make sense to compare with the offline world because this wouldn’t be able to happen outside the internet.
In response the CEO is attempting to force them into a different context where he once again has power. The author recognizes this and therefore refuses.
This is psychotic behavior.
There’s a huge spectrum between NY Times writing a sourced article about a powerful business magnate and someone disparaging an SMB owner on their blog. If I took the posts and emails of someone I knew in my life and posted them online, I would probably get a call from the police.
I feel that Vlad is justified, even if I personally would've just considered it to be a lost cause and just kept receipts in case it became necessary to publicly respond, similar to how the Apollo dev released receipts when Reddit tried to make him out to be in the wrong.
I've also had a similar discussion with Vlad on comments here, he definitely doesn't try to view things from other people's perspectives.
https://d-shoot.net/files/kagiemails.txt
Read them yourself, but to me they look like the emails of a persistent salesman. They were remarkable only in that they provide more excuses than concrete responses.
> Not even Google ever printed 20k tshirts to give away for free.
For a couple of my university years I had nothing but free Google t-shirts. They were throwing so much of this crap around that my closet was halfway to 20k. I only lamented they never gave away Google trousers or briefs.
They have a fair shot at competing with Google on quality of search and they should focus on that. If they think they can complete on AI, email or swag - good luck, and I hope you have a good money printer.
I actually stumbled across the AI stuff being turned off by default yesterday when I got curious and was poking around the feature request forum. It was explicitly because a lot of people hate it for moral/ethical reasons. A lot of the comments in the replies are specifically about the AI stuff in spite of it being disabled by default.
Most of this seems fine for a startup?
>Yes, hello so called prince of Nigeria. I have no interest in a discussion about the intricate court politics of Nigeria or its Byzantine inheritance rules. As you can see from my blog post it is entirely unlikely you would ever gain the throne even with my $2,000 wire transfer.
The only thing I take away from that is I'm very happy I don't know either of them and am never likely to.
I understand not wanting to engage in a conversation about a product you don't care about, but after collecting so much information and writing a lengthy blog post about it, that is a different story. In my eyes, the author wrote a hit piece largely based on personal grudges, and then wanted to avoid any kind of responsibility.
And from my point of view, a lot of the financial stuff "makes sense". This is a small startup, probably with little business experience, and it shows. But why make it look like they are doing evil because of small, negligible mistakes?
Narcissism.
> I may not have spelled this out explicitly in my previous reply but I will do so here: I am not interested in getting more replies from you on this subject. Declining a call does not mean I want you to argue with me about Kagi in email either. I do not trust you, personally, either and do not want to have a conversation about that. And for the record, I read that blog page already. If you had read my own blog post, you would notice that I link to it.
replying with a 1100 word long email is a mood.
What mood?
> I may not have spelled this out explicitly in my previous reply but I will do so here: I am not interested in getting more replies from you on this subject. Declining a call does not mean I want you to argue with me about Kagi in email either. I do not trust you, personally, either and do not want to have a conversation about that. And for the record, I read that blog page already. If you had read my own blog post, you would notice that I link to it.
If they don’t want to talk, just don’t respond.
The author also cross posted their blog to multiple social media platforms, which I assume means they wanted it to get attention. But then when the CEO does see it and offers some explanations they get mad that the CEO “vomited out” a reply that they didn’t want? I’m sorry, but the CEO of Kagi definitely sounds like the reasonable one here, thanks for linking this thread.
Like, these days you do not know when you email someone if they reply to you, or if they will post screenshots of your entire conversation to social media showing how utterly disgusted they are because you dared talk to them.
Have these people forgot about how strangers in real life behave and communicate?
I miss talking with the average idiot from the 2000s internet.
Yes, my impression as well. (I have never used Kagi but have considered trying it.)
Among the other things, the blog author approvingly put up a screenshot with someone insisting on seeing the entire world through their own political views and demanding others do so as well. ("Actually, the word 'politics' means 'everything', and also I'm right and everyone else is wrong.") As the meme goes, they need to touch grass.
Perhaps Vlad is a little excessively enthusiastic and protective of his baby. But then you don't do something frankly crazy like start a new search engine from scratch in 2023 without being a little bit off. If we actually want a viable alternative to the advertising-funded search monopolies, we've got to be tolerant of some personality quirks.
And perhaps the T-shirt gambit is a poor use of limited resources. But have any of the startups that ended up making it big not make a few poor investments on the way up? I'll forgive it.
Meanwhile, Vlad's response does spell out several ways in which this lori exaggerated or misinterpreted things. Which of course are not acknowledged or responded to at all, despite lori's self-important tone. If you want to take your ball and go home because somebody doesn't take your concerns seriously, well you can, but don't expect me to follow you.
IMO, Vlad would have been better-off making his response his own blog post somewhere rather than an e-mail exchange. But eh, at least it's out there.
That's said, I've been a customer for a while and the t-shirt debacle is one of the dumbest things I've seen a small company do. Even if you try and call it marketing cost (no name on the shirt makes that hard), there's no way it was the most efficient use of money for marketing.
And setting up infrastructure for it wreaks of "I'm bored with search let's do t-shirts." it completes goes against "do one thing really well" and just seems like a waste. If I were one of those investors and my money got spent on that I'd be really upset.
If they thought it'd help, absolutely. Facebook has a long history of doling out swag; I've got a free Oculus sitting downstairs.
It remains the worst customer retention pitch I've ever seen.
Maybe I'm being pedantic but Kagi doesn't own a T-shirt factory and presenting it as such is a bad faith argument that does make me question the author. They very clearly point out that they worked with a print shop in Serbia to make the shirts.
- Businesses that’ll eventually need to collect sales tax should add sales tax automation software to their stack from the beginning (even if they don’t expect to hit the threshold in a given jurisdiction anytime soon). Compliance can get really complicated really quickly -- and what founder wakes up and prefers doing taxes over actually building their product?
- Avalara is probably the most well-known one (Box uses them) but they have an implementation fee that we couldn’t stomach as a bootstrapped business https://www.avalara.com/us/en/index.html
- We ended up going with a fellow startup called Kintsugi because they had a free option that connects to Stripe and didn’t have an implementation fee. We probably won’t even hit any thresholds to need to file until 2025 (cross your fingers it’s earlier!!) but they said they’ll email us when we do, here’s their page https://trykintsugi.com/
- Anything related to compliance makes me anxious so this is great because I don’t even have to think about it until I actually need to, though hopefully not under the same circumstances as Kagi :)
I think this is only going to become more apparent once AI-generated content takes over the web.
One thing I want to look into is ranking algorithms based on individual engagement. So, if you save lots of stories from a site, it ranks higher. If lots of people save stories from a site, it also ranks higher, ect
2. Zero tolerance to bots
3. The service is not free
4. Individuals have very little impact on the rankings of other users, so you need to pay for a lot of bots
I believe that the true problem with bots on, say, twitter, is that they have perverse incentives to 'boost engagement' and whatever by allowing the bots to run rampant.
That may fix things on its own ;)
Now would I pay to see what other people 'engage' with? I'd probably associate it with the likes of facebook/twitter/other social networking crap and I'd just move along...
All information well organized without ads and junk. Like a super glorified Wikipedia with excellent search to dig exactly what you want.
Kagi is for a subset of the internet and specifically for the part that has content. The good parts of cyberspace if you like. OP seems to be looking for something bigger like someone they can trust to replace Google and save the internet as well. For that search I say good luck sailor
(see, that is the good thing in Kagi too - you can downvote ;)
I'll just say the quiet part out-loud: expecting people to pay $10+/month for a search engine is a pipe-dream that rules out 95-98% of the world population. People buy food with that money, they pay rent, they live lives that aren't tethered to a search engine in a meaningful way. Google "wins" their traffic because they don't care, and every bit of friction in-between them and their content is extra work. Kagi's payment-upfront mentality is unrealistic for everyone except the most well-paid Bay Area engineers.
That's not to say I don't understand the "avoid ads at all costs" concept. I do oppose to using anti-advertising sentiments as a populist rallying cry so people will line up at your Search SaaS kiosk and pay you whatever you ask for. At this point you really might as well just invest in your own Searx instance, it's plenty cheaper. And you can't even "dropbox comment" me since there have been third-parties providing search for free since before HN was a website.
So what? Why do you get upset about it, when nobody is forcing you to buy it? Most people will not be interested in paying for search, whether they can afford it or not. That's just what a niche product is, most people will not be interested. What I produce in my job is certainly uninteresting for 95-98% of the world population, and the same is probably true for your job.
> Kagi's payment-upfront mentality is unrealistic for everyone except the most well-paid Bay Area engineers.
It's ten dollars.
> At this point you really might as well just invest in your own Searx instance, it's plenty cheaper.
Yes, that might be a good solution for 95-98% of the world population.
Because this isn't a solution. Kagi doesn't save people from advertising, it creates a premium workaround and sells it at an arbitrary price per-customer. It's software-as-a-service, a SAAS, built more for the 1,000 true fans rather than the 100,000,000,000 clueless web users. That's just another business - perhaps a kinder and more transparent business - but a sinkhole of regressive moneygrubbing all the same.
> It's ten dollars.
Which is ten dollars more (per month!) than most people pay for a search engine. If you're the sort of person that just flippantly subscribes to that, then yes, you have lost track of the value of a dollar in my eyes. Like I said - you can host your own search engine and pay for your own top-level domain at that kinda price. It's absurd, I'd protest it on-principle even if I was upset with my current search provider.
There's room for this sort of startup in the world, but they've already lost if they don't offer a free tier. Google will hoover up their potential customers like nobody's business until they take the 98% seriously.
It is a business, what did you think? That's why they charge money for their service. Like millions of other businesses, they will never get any significant part of the world's population as users. Why is that a problem to you?
I also think this product might be a bit too late. GPT4 has been out for over a year now, and it's changed how I look for answers. I tried FastGPT but like the author I found it lacking. As it stands, Perplexity feels more like the future of search than Kagi.
I’ll be curious if you come back to Kagi in a couple months too.
Being able to either outright block/pin sites, or only lower/raise them in your search results, made a difference for me after a few days of searching.
I do hope the future of search will include being able to use natural language, but also still the more precise '"This" +"That"'.
In any case, more competition is a good thing, the lack of it is what got us into this mess in the first place.
I don't care about Orion and Email, but what I'm getting right now in terms of search experience is definitely worth the cost.
Didn't even know that existed, thanks for sharing.
The thing that made me cancel my subscription was one specific interaction.
One day I was trying to buy tickets to a podcast tour, the sales for tickets was set to open at a specific time and I was searching for the purchase page at the moment of opening. I frantically searched "$SHOW_NAME $CITY tickets", the first search failed to bring relevant results. I tried "$SHOW_NAME $CITY tickets $YEAR", nothing.
I tried many searches for about a minute along these lines and thought maybe their site just wasn't public and I needed a specific link. Then I typed my original "$SHOW_NAME $CITY tickets" query into bing and got the exact correct webpage on the first try.
Bought the tickets I wanted and immediately cancelled my subscription to Kagi.
This article comes across so unhinged it almost works as an advertisement, except for the founder dismissing privacy issues...
I'm happy to hear Kagi are creating an e-mail service though, I've been looking to get away from Microsoft 365 since I'm not really using the meat of it. I hope they allow multiple aliases per users and perhaps add a masking service as well.
Kagi: "The process from here involves setting up a business entity in Germany, so we can import the t-shirts, store them in a warehouse, connect inventory logistics and ship them all over the world. This includes building a website and connecting it to a back-end database. So, we basically ended up owning a merch production operation end-to-end, just so that we could ensure premium quality of these t-shirts!
Now, you may ask, why did we go through all this trouble and allocate nearly a third of our investor-raised funds to produce and freely distribute 20,000 t-shirts?
We would not be here without our early adopters (you!) and we deemed it important to pause, reflect and show gratitude.
We acknowledge that our journey is a marathon, not a sprint. With a long road ahead, supporting our member community is both rewarding and meaningful.
Simply put, wearing the Doggo t-shirt is an incredibly awesome experience."
Kagi Search is at the very least intriguing, though I honestly didn't find the results very impressive; they seemed alright, but nothing spectacular. The thing that is frustrating is, Google has a massive index, but searching it is an exercise in frustration because it feels like it is basically rewriting your query. Even using "" and + no longer seems to be good at ensuring certain things appear in the results, and so I sometimes try, in desperation, to simply repeat the term I want to emphasize multiple times in the query, which finally sometimes allows me to find things I already know exists. God forbid you wanted to find something you didn't know existed, because in that case, you might never realize Google is fucking up what you're looking for; it has the answer, but it's hidden in a sea of Google-funded blogspam. What a mess.
Will there ever again be a profitable search engine that works as well as Google used to? The answer might be no, and this bums me out.
> And he is very, very much the type that believes "not everything is political"
Well, at least we agree on one thing, I have always felt the "everything is political" angle was one of those semantic technicalities, kind of like saying "actually, the glass is always full, just sometimes it's full of air". The lack of a well-defined boundary between "political" issues and non-political issues should not be used as an excuse to drive politics and politically-charged discussion into otherwise rather mundane and apolitical things. I suppose it's not really that important, but this is one of those Internet-era brainrot issues I dislike most. Of course, maybe this is actually trying to make a more nuanced point, but it being phrased like this activated my "uhm, actually" response impulse.
Also, anyone who can say with a straight face that their preferred political party is aligned with the truth, while the opposition is aligned with lies loses credibility. People lie about anything whenever it suits them, and in politics and the news that is rather often.
It just so happens that they like what the one side is saying and not what the others are saying. So the one side must be good and telling the truth, and the other is bad and full of liars.
It's hard enough in a small startup to prevent CEO "commentary-driven-development" , let alone have their random thoughts driving investment insight and user acquisition/attrition.
It’s good interacting with the real people that make software.
IMO, The fact that it’s so detached from the customer is part of why MBAs fit in to leadership so nicely.
None of the customers see it coming, because they don’t interact with employees.
Seems like a feature, not a bug. The only politics I care about from my tech products is whether they will collect/sell my personal data or not. Too many tech companies use a veneer of support for various political causes to take attention away from their own misdeeds. At this point it's a red flag when a company talks more about their politics than about their product.
Thus the quotes at the end are quite bad:
> people who really need anonymity are very rare. probably less than a 100 in the entire world. definitely not typical hKagi users (edited)
> unless they are criminals, in which case we don't care they don't have full anonymity (nor we want them as customers)
Women seeking abortions in the US weren't criminals two years ago and now, in some places, they are. I don't think he's thought about this position very carefully.
The only issue that has been brought up here that concerns me is privacy, and the important thing there is that whether search history data is stored and can be linked to and account. Some things like personalisation features require it, but what about the rest? It does concern me that I have to take their word for it.
At the end of the day, Kagi was a way for me to filter out certain sites and raise/lower others. To be honest I’ve considered turning off the ranking modifications, I often have to scroll to find the business I’m searching for’s website because I’ve up-ranked SO/GitHub/HN and the like. It’s more frustrating than useful. I wish there was a way to up-rank the “definitive website for a brand” when I search.
Furthermore, the lack of local results is painful. I just have to go to Google to find the restaurant/business in my town since Kagi seems unable/unwilling to do that. And in that vein, all my !bangs seems to have disappeared which is frustrating, even more so since on mobile it won’t save new ones (last time I tried) and manually going to Google gets redirected to Kagi due to the way their extension (has to?) works.
I’ve been paying for a while now and while the AI doesn’t bother me (I don’t use it) the Brave stuff turns me off massively. I don’t know how I’ll decide but I found this post (and the emails the founder sent) very informative.
It turns out that even though I can't stand the number of ads, Google is still much better at getting me an answer quickly (usually with the quick answer modules).
I was also surprised at the number of times Kagi came up with 0 search results, and while one of the draws for me was to have higher quality results instead of quantity, I still found a _ton_ of results for AI generated crappy top-10-list sites trying to sell me something.
Love the idea, and will probably check back in from time to time, but so far the execution just isn't there for me.
Gathering feedback good. But getting involved in philosophical discussion or how to run the company looks like a bad idea.
This does invite trouble but interacting with users of the product I am building is also the only way I know how to do it and is keeping me sane. Not to mention it helps build a great product, as users probably 'built' half of it with feedback.
I never thought that talking too much with the customers can be bad but it also may be true that full openness approach becomes a burden at some point and that it would be healthier to separate from it a bit.
It seems pointless. I can sort of see why Microsoft would do it, but that's Microsoft wanting a modern browser for their operating system, not DuckDuckGo, Kagi or Ecosia wanting a browser for their search box.
Why this is pretty much just a weird rant about Kagi, I do agree with the questioning of the investment in maintaining a browser.
The thing is they can't control the stability of alternative methods on Safari. It's up to Apple.
- Safari + Kagi: you just need a default browser and a non-default search
- Orion + Kagi: you just need a non-default browser and a default search
- Firefox + Kagi: you need a non-default browser and a non-default search
It might not look that much difference but in terms of thousands users, any reduction for users' steps to do, it is better.
Honestly, this gave me the impression that Kagi is a move fast and break shit kind of company, with Vlad as a benevolent dictator type. Finances are often (I think) a disaster at these kinds of companies that gets fixed later. The t-shirt thing is sus but those shirts look really cool and I might buy one. Vlad being flippant about getting sold is better than lying about "nothing will change; we will find a buyer that agrees with our values" while looking for the highest bidder, values be damned. The email thing makes sense technically (they're basically trying to be a Google alternative) and seems like too much, but, hey, it's Vlad's money (I believe he bootstrapped this? If so, the investments they got were probably more donations than investments) and if the devs want to do it, more power to them!
Vlad is "my way or the highway" unless you convince him that the highway is actually better. I've seen this happen with many feature requests, like them improving WebAuthN support or integrating support for Apple's iCloud Keychain extension, both in the Orion browser.
Also, while Kagi seems to have pivoted from an AI answering service to search, it seems like a really logical pivot and comments like this:
> 1 AI should be used to the extent that it enhances our humanity, not diminish it (AI should be used to support users, not replace them)
give me the impression that they're at least trying to be responsible with this tech.
Everything in this article is mildly concerning, but the alternative (going back to Google Search and embracing a mono-browser future) is so much worse. I don't regret supporting their mission at this time.
Making a bunch of free shirts is something DuckDuckGo did too, I have one. Is it a bit weird they incorporated their own factory for it? Yeah, but I guess I just figured that that was part of their marketing budget. How many cool products have we seen die just because their marketing was crappier than their competitors?
I do think that building their own browser is a waste of resources, but I think that the AI tools also are kind of marketing. People utilizing the universal summarizer gets the name out there, and allows people to see the Kagi branding more frequently. Are they spreading themselves too thin? Maybe, but simultaneously I do feel like the second that someone has to reach for Google to do anything, that's going to whittle away at their potential market share. I think the fear is that people are going to say "if I have to do X using Google anyway, why should I stick with Kagi, especially since Google is free?"
It's tough to say, I've been a paying Kagi user for about a year, I like it, it's a product that I actually think is better than vanilla Google. I really want them to succeed, so maybe I'm viewing what they are doing with rose-colored glasses, but most of the complaints in this blog post didn't seem completely horrible to me.
And it's not just recycled google results. It's google's API, which means that it brings back into the light what google was best at before they decided to monetize it with echo chamber algos and toxic ads. I can find it in my heart hate Alphabet and yet pay to mainline their best product with a sprinkle of other vitamins and minerals like Wolfram.
I share the concerns about wasting time and money on the other silly side projects (t-shirts, email, maps..), but as long as they don't negatively affect the search product or sink the entire company I don't truly care.
I do not like that their promise of privacy relies entirely on aligned incentives and trust, but it is no worse than the alternatives that all rely on ad-tech money.
Discord is an excellent place to collaborate. It’s easy to use for non-techies. The level of community is consistently high, better than old Reddit imo. Being realtime it’s conducive to ad hoc chatter.
Like others, I think that they had to raise money is a worrying signal they are not sustaining themselves. However, that’s only a signal.
I may not think the CEO is a particularly nice guy (though he was not too pleasant when responding to tickets, which yes he apparently does himself (unless it’s a shared account?), he was not rude either), but realistically can I demand that from him? Push comes to shove, can’t really claim I’m extremely nice either, and I have achieved less.
Regarding the use of ML or Kagi being originally an AI company, I don’t think it necessarily condemns the project.
Kagi Email drama seems ridiculous, but is very new to me.
Unless there are some further bad revelations about the company or the CEO, I will reserve my judgement for now…
"why pay for shirts or advertising when we can own it?"
Anyway, that's how I felt it happened. Build over buy. Siphoning is an interesting thought
It's mind blowing. The earliest internet ventures were selling shirts that somebody else made for huge margins.
Why take on the costs in-house, lol. Either way: by making it yourself or dealing with problematic suppliers, it's not worth the hassle.
Shirts aren't their business and for those still in it, the margins are razor thin. Madness.
Useless snark aside, if the Kagi team wanted more "money directly into their own pockets" they could just raise their salaries. If their product is comparable to Google their salaries could be as well.
And they'd get the money for those massive salaries from what magic money tree? They're at best barely profitable, revenue is probably around $2m, and they only raised $670k in funding. This isn't a particularly money making enterprise.
Personally Identifiable Information includes anything that can be used to uncover a person's identity. An email address is PII, as it can be used to identify which person their data relates to - and they do have data, at the very least a user account and settings but likely also logs.
Full names, phone numbers or IP addresses are also PII - if you have a server log with source addresses, that's PII under both GDPR and CCPA. Why you have it, and whether I can take steps to hide my identity is no excuse - you need to follow the legal process for PII under GDPR and CCPA, need have data controllers in place and ways for any individual (registered or not!) to make the appropriate data requests and removal requests as applicable.
To determine whether a natural person is identifiable, account should be taken of all the means reasonably likely to be used, such as singling out, either by the controller or by another person to identify the natural person directly or indirectly.
There are people who argue that just the name of a person is PII and they are wrong.
It's very easy for a name to be PII. I'm quite certain mine is unique, due to hyphenating when I got married.
Whilst the response from OP is so obviously wrong and confusing that it's likely to be a troll and not worth engaging with, it's worth clarifying to anyone reading this thread that email addresses most certainly do qualify as personal data under GDPR. GDPR very clearly states what personal data is (see https://gdpr.eu/eu-gdpr-personal-data/ and https://gdpr-info.eu/issues/personal-data/) and that storing or processing of this data necessitates the need to comply with the requirements of the GDPR (particularly the rights detailed under https://gdpr-info.eu/chapter-3/).
For the purposes of this conversation, an email address is personal data, operating in the EU (and additionally, by way of carried-over legislation, the UK) means complying with the GDPR, and therefore Kagi need to provide mechanisms by which people covered by the legislation can enforce the rights afforded them within it.
(GDPR also doesn't use the term "PII", merely just "personal data" and goes on to detail what this means in terms of identification, which might add to the confusion in OPs original message).
It's in your database. That's inherently a connection to something else.
> ‘personal data’ means any information relating to an identified or identifiable natural person (‘data subject’); an identifiable natural person is one who can be identified, directly or indirectly, in particular by reference to an identifier such as a name, [...]
https://gdpr.eu/article-4-definitions/
A name is PII, as stated by the definitions in the GDPR.
What is there to misunderstand?
"This Regulation applies to the processing of personal data wholly or partly by automated means and to the processing other than by automated means of personal data which form part of a filing system or are intended to form part of a filing system."
A name or email in Kagi's database is very clearly subject to GDPR. A note on your desk may not be; not because the name isn't PII, but because not all PII is in a protected context.
You're incorrectly mixing up "it's not PII" with "it's not subject to GDPR". It's still PII even if you're not legally required to protect it in a specific scenario; I can, for example, tell random people about my wife's very unique medical conditions, but her hospital cannot.
Quoting GDPR Article 4, point 1 (https://gdpr.eu/article-4-definitions/):
> ‘Personal data’ means any information relating to an identified or identifiable natural person (‘data subject’); an identifiable natural person is one who can be identified, directly or indirectly, in particular by reference to an identifier such as a name, an identification number, location data, an online identifier or to one or more factors specific to the physical, physiological, genetic, mental, economic, cultural or social identity of that natural person.
And Recital 30 (https://gdpr.eu/recital-30-online-identifiers-for-profiling-...), which gives some more examples of identifiable information such as IP addresses:
> Natural persons may be associated with online identifiers provided by their devices, applications, tools and protocols, such as internet protocol addresses, cookie identifiers or other identifiers such as radio frequency identification tags. This may leave traces which, in particular when combined with unique identifiers and other information received by the servers, may be used to create profiles of the natural persons and identify them.
There is also the quote from the Danish Datatilsyn (https://www.datatilsynet.dk/english/fundamental-concepts-/wh...) also makes some more examples and explicitly highlight that is is PII even when it must first be combined with other data:
> Personal data may, for example, include information on name, address, e-mail address, personal identification number, registration number, photo, fingerprints, diagnostics, biological material, when it is possible to identify a person from the data or in combination with other data. It is said that the information is “personally identifiable”.
---
An email address is obvious PII because it is a globally unique identifier representing a way of contacting a specific person. You can find the name of the owner separately and correlate it with your stored data, thus identifying the person.
Even if you store nothing else, the email reveals that you have an association with the user, but you most likely also have data that ties activity to the email such as the user logging in and using your service in any way or form.
> There are people who argue that just the name of a person is PII and they are wrong.
A person's name is the most obvious case of personally identifiable information.
And whatever the heck is going on with all the other stuff. There's no way one should stretch oneself this thin.
I’ve also already stopped using Kagi on mobile, because they made the experience of editing the search field terrible with some JS. In particular moving the text cursor in the search field in Kagi on mobile Safari is an exercise in frustration that’s not replicated by any other text field in any other search engine, website or anything at all on iOS.
This confirmation about their core focus being LLM stuff seals the deal for me and I’m cancelling my subscription. I fully agree with this sentiment from TFA:
> They just don't want to admit to being an AI company anymore. Frankly, it's not something I want to pay them to keep developing. It's something I want less of out in the world.
I still do like the product they offer, though. It'll be difficult to give up their bullshit result filtering. I also cancelled my ChatGPT plan because I could use GPT4 through Kagi. They also provide access to Gemeni, Mistral, and Claude. Probably actually an unsustainable value.
Can you clarify what makes you say that?
Orion development is very active. Here is Orion's changelog in case you missed it:
I also don't want to be overly critical, and I am a paying Kagi user, love it and will eventually try to use the family plan, but maybe I'm just not at the point where I have time to deal with glitchy software anymore.
I use it as a daily driver and many other people do and while there are many issues still (we have over 2000 open issues on oriofeedback.org) Orion has never been better and I encourage you try it again.
Why not focus on doing one thing properly?
Although you seem to obviously be attracting enough money so it's up to you how to spend/burn it :)
I think it's a crazy ambitious bet but I can see why you're making it.
If it's deeper than that, I'd recommend adding a summary. It's a big ask to expect strangers to read a ranting wall of text.
> Is the search good? I mean...it's not really much better than any other search
It is absolutely fantastic as a search engine. The few times I find myself on DDG or Google I am reminded how terrible the alternatives are. For this alone I am happy to pay for however long as possible.
In terms of letting me find what I'm looking for, Kagi is the best search engine I've tried. At worst, it's no less "private" than the alternatives.
The AI stuff is irrelevant to me until it starts degrading the search results I get. I just ignore their AI-related features. No big deal, and certainly no worse than others.
All the other criticisms (t-shirts, etc.) don't matter at all to me. What matters is search quality.
The search is generally good enough, often really good, and the killer feature is FOREVER hiding results from low grade sites and boosting results from sites you respect, without the tonne of commercial rubbish that Google feeds into results.
Where I see an area to improve is image search - Google is usually better and it's usually just laziness that I don't do all image searches in Google.
I wish Kagi luck and certainly hope to continue to be a customer.
I don't even like the dog mascot.
I also refuse to use anything from Brave for this reason.
This is not about separating a person's work from their political views, it's about not giving more power to people you don't want to have more power. If he is going to use money to campaign against LGBT people, then I'll do my best not to give him money, and I think you should too.
It's not that hard to prioritize companies that are aligned with your values. You don't need to start living in a cave or self-flagellate. Just look around to see if there is a suitable alternative.
Then yes, I use Google for my search engine. They do shady stuff, but they never campaigned against same-sex marriage, and none of the alternatives have given me good enough results and an acceptable political alignment for me to switch. Brave certainly doesn't seem much better, with unacceptable political stances and worse search results. DDG is getting there, but the search results are still not good enough to replace Google for me.
I don't mind the financial misplay(?), the possibly unsustainable diversification, the lil bit of corporate hypocrisy, or the AI - they're endeavours that can't be avoided while trying to make a compaany profitable - as long as they do what they can do best, WHILE maintaining the stance & spirit taken for forming the company.
For instance, he claimed both that a search engine shouldn't judge someone for searching how to commit suicide, because they might just want the equivalent of legal euthanasia in a place where it's unavailable and a global search engine should be value neutral for it can't possibly know all cultural norms, but also said "it wouldn't work" as a backup argument that was promptly ignored when I provided a study contrary to it.
I ended up feeling like Vlad has a lot of implicit beliefs that are rooted in free speech absolutism, but will rationalize via other unfalsifiable arguments (like scope creep / too many unknowns) while appearing like he'd be open to be convinced. I still use Kagi, but I'm ready to jump ship at the first opportunity.
If there were a way for an FSF-adjacent company to have a user-owned business model that could both:
a) maintain fidelity with FSF goals and ethics
b) also do the hard-nosed business decisions that are needful.
As a decades-long FSF member, I'd cheerfully support subscribing to a set of email/search services, assuming technical heft.
The two mindsets seem hard to harmonize.
I don't care about closed-sourced browser nor their AI offering
Is there a common thread amongst search engine founders? The guy from you.com also gives interesting vibes. Here he is with 10 shill accounts. https://news.ycombinator.com/item?id=31344999
Edit: I think Vlad is a true believer in what he preaches. I detect no subterfuge in him. But he takes negative comments personally as if people have a duty to love his vision and product. A true believer but seems like that would be a tough trait to have when you cater to a very opinionated audience.
> Our payment processor does, and you can ask Stripe for that
- Kagi CEO
> The data subject shall have the right to obtain from the controller (...) access to the personal data (...)
- Art 15 GDPR
> Controllers are responsible for complying with SARs, not processors.
- ICO https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-re...
Kagi = controller / Stripe = processor
So no Vlad, that is not how the GDPR works.
That said, one of the things that kept me from using Kagi much is their focus on A.I., which has been staring us in the face since Day 0. I hate A.I. Anything connected with A.I. carries a certain stain that is hard to articulate, but basically immediately makes me feel I have some insight into the psychology of the people behind it, and I do not like what I see.
I also think that a lot of people assumed that Kagi is privacy respecting without much evidence, it was never part of their pitch that I remember.
For those reasons I've just never used the service. This author's axes to grind beyond that don't concern me too much, but those two things are total deal breakers. Which is sad.
More variety in the search space is still good from my perspective. I still hope that Kagi succeed.
Also, can't wait to get my tshirt! (closed beta user here)
I don't get why random punditry is of any relevance here, or why Vlad felt the need to respond. He needs a PR person, I think
Could have fooled me!
User scripts are doable on Firefox for Android.
For iOS, somewhat ironically, I think you'd need Orion.
But in all seriousness, I’ve been a subscriber ever since they started and I’m an ultimate subscriber still, and I’d be sad if they went bankrupt due to mismanagement of the funds.
There is no reason for Kagi to remain “pure” and avoid AI features as suggested by hardcore AI haters. I am not a fan of AI hype either, but I am pleased to see that Kagi has integrated some moderate capabilities such as the summarizer and search-based generation, which are natural extensions of a modern search engine. (I do hope they improve the expert mode soon, as it is currently far inferior to Perplexity, but that does not invalidate the general point.)
Email-based account management may not be perfect from a privacy perspective, but registering with a privacy email alias has mostly resolved my concerns. As for GDPR, let’s not pretend that it is disproportionately burdensome for startups. I value the way a company operates much more than the privacy theatres (banners, opt-outs, legaleses) enforced by GDPR.
Other criticisms regarding operational details range from nitpicking to trivial. I do hope that the founder was less insistent on arguing with and lecturing zealous users like the author.
Search just works, 90%+ of my searches are on kagi. Much better than google, bing, ddg, etc. Worth the $10.
I do use fastgpt and the summarizer sometimes. As with any of these AI tools, you have to get a feel for what suits it and when to use it.
The unbiased review idea _sounds like_ it would be ultimately fruitless as the requirements get philosophical, but it may very well lead to more useful tools in the process.
The GDPR perspective is unfortunate. I'm willing to wait and see if they eventually accept it.
Overall, I like their focus on user experience, customization, fast and light websites, and search quality. Sure it might not as objective as it's portrayed, but it is giving me great tools _today_.
> Like most search now Kagi has chosen to include Instant Answers that are AI generated, which means they're often wrong, as well as a "Universal Summarizer" tool, that again is more of the same old AI bullshit.
I agree that Kagi should focus their efforts on delivering the best possible search experience (the image search is horrendously slow at times). But as for the above-mentioned AI tools, I love and use daily. For a short question style query, most of the time, the quick answer is all I need. Universal Summarizer is excellent at summarizing the uncountable YouTube videos that my friends send to me.
> At one point someone suggested the idea that searching for suicide-related terms should bring up a helpline, and he rejected that idea because it would be "biased" (I guess towards not wanting people to kill themselves).
Why is the so hard to accept for some people that not everything should push a message? If I want to know about a controversial topic, I need results about it, not a lecture. Their responsibility is to serve me good results.
I view Fastmail in a similar manner, but the difference with them is they have a real business market for those wanting an alternative to Google or Microsoft.
Personally I don't use Kagi for 2 reasons: It's not significantly better than what I already have (SearXNG private instance) and I don't really subscribe to the "everything must be paid on the internet". I'm much more comfortable running some dockers and putting a privacy barrier between myself and google, as I do for my searches, for my phones, computers etc. Of course this kind of tech solution is not an option for everyone, but I don't care about everyone :) $10 is a lot of money for me and I rather spend that on something I control like 2 VPSes.
Also I think that once things go commercial eventually they will reach a scale where they will want our data too. After all, data is like free money. A business will never say no to free money. By due diligence law they're not even allowed to. Kagi has that luxury now, but once they become big and own half a northern californian town for their campus they no longer will. That's the idea of enshittification.
Additionally, by running a private instance you have effectively given all these companies a spotlight onto YOU. Only your searches are going to come from that instance and these companies already have enough information they can figure out WHO is making that search pretty much right after you click through the first result.
Just from the available public instances you can quickly see that other search engines may block your requests for one reason or another but your server will be constantly retrying to query them from time to time. Last thing I want is to find my IP addresses on some sort of "naughty list" because I wasn't honoring some 403 error my server was getting every time I searched something.
There is nothing that says other search engines won't wise up and figure out how to stop searx from "abusing" their search functions so it almost feels like the only sure way to ensure search privacy is to run your own crawler, for better or worse.
And of course I block cookies and ads heavily, bypass paywalls, the works. I'm sure I'm pretty trackable anyway with so many addons. Also I'm probably one out of 10 that uses my specific OS/browser combo. So with fingerprinting they have got me anyway.
I run it on a VPS so I don't really care, that one is not used for exit traffic of any other kind except an IRC bouncer. And if I really need to I can just switch to another IP :P
It is quite strange that they are doing all these other things that I basically don’t use at all though. Classic entrepreneur chase shiny objects and get bored with the core idea. In their defense though, people do believe that search is going towards an ai future.
But I do want to say that Kagi Search is really awesome and I hope it works and I won’t go back.
Speaking about their whole business, I think three things left me a negative impression:
- the tshirts were really unnecessary. I didn’t understand that. I am not sure the world needed more trash being produced and for sure it was not a good use of their money.
- I think AI as a tool has a place in their offering (Quick Answer, Summarizer). I don’t think the Assistant stuff makes sense for a search engine.
- the apparent lack of care for privacy that appears in the quotes in the blog posts are not good and I hope Vlad changes his mind and addresses that properly. Everyone needs privacy. Moreover, GDPR is no joke and it should be followed properly.
Kagi maps seems to just be Apple maps otherwise I don't see why they would put Apple maps logo in the bottom left corner.
I'd love to see more focus on their own search engine/results, as well as technical means of ensuring anonymity/unlinkability within their infrastructure, but they're well worth the $250/yr for a pro subscription today.
There's some kind of psychological instinct that makes some people think that they are owed something and have some kind of personal relation to somebody famous or in this case a company, a kind of familiarity that is just one way. The author of course wouldn't direct this kind of attention towards Google or any other huge company, because they understand that there is no relationship between him and them. But now when it's a small company, there is a short circuit. Just like stalkers usually direct their attention to female performers, and not to gangster rappers or a rock band.
Of course there will be nothing that Kagi as a company, or the people behind Kagi could do to please the author. When in "hater mode", exactly everything the other part does or says will be turned and twisted into something bad. Just read the e-mail exchange that was posted.
With that said, Kagi is still the best search engine around and if they someday won't be, it's as easy as unsubscribing.
I think Kagi have a lot of things to answer for, in particular, a large tax bill, possible GDPR violations, and potentially a future inability to pay their hosting bill in t-shirts.
How could someone come up with this idea and think it was good? For a search engine whose user base is so interested in better quality results they're willing to pay for them?
I get it, chips are stacked against them, but something as critically well understood as searching the web just HAS to work as expected, because the free alternatives have a very very good UX.
(Agreed about manually going to kagi.com not happening, etc.)
... What?
Has this person ever visited a review site? I mean, it varies fairly dramatically, and there are some niches where 3/5, 5/10 etc has a fairly defined meaning (for instance, for TV reviewers it means "meh", for gadget reviewers it means "this will catch fire as soon as you plug it in"), but really, I mean, what?
For example, Kagi very clearly does not own a t-shirt factory, and this worst-faith take makes me distrust the entire post.
Yes. The part where the CEO claims that of all the people seeking anonymity that less than 100 of them aren’t criminals o_O
Check out the last discord screenshot.
I'm a happy, paid Kagi subscriber and have been one for a long time, but I've been uneasy for a while about their lack of focus (the T-shirt fiasco being only the latest example) and the post demonstrates clearly that the issue is systemic. You're trying to compete in search against Google and Microsoft with 12 people! Stop doing irrelevant bullshit that's no going to improve your bottom line!
He was successful about building a community around the store but not successful at the paperwork so it turned out we had not paid the sales tax for a few years which led the state to put up signs in front of the store, thankfully he was able to scrape up the money. Boy it was a near death experience.
(2) If you are working on search in 2024 and you are serious you’re going to be using A.I.
Pretty sure Marginalia doesn't use AI and to the best of my knowledge Viktor hasn't written about plans to do so. But maybe he's not serious because it sure seems like he's having fun!
PS: Lovely username =)
(2) One of the most depressing experiences of my life was reading through the first ten years or so of TREC conferences looking for something useful to improve the search engines I was building. Eventually I found a volume that revealed the handful of useful results that they got in the first ten years (here https://mitpress.mit.edu/9780262220736/trec/)
Advances in search quality are rare and come along about once a decade; BM25 was such an advance, on paper. Even though BM25 is in Elasticsearch and a lot of other products very few people are taking advantage of it because they don't want to do the parametric tuning it requires to get superior results.
is a similar once-in-decade advance that actually works out of the box with relatively little tuning. It doesn't address all the issues of search and should be integrated with more traditional search, but if you are building a search engine in 2024 you can expect to wait another 10 years for another advance like that.
(3) Marginalia particularly interests me because it is a small collection and the problems of search over a small collection are very different from those over a large collection. Gerald Salton started IR research with a deck of punch cards and he thought 80 documents was a lot and with 80 documents you are going to be very concerned about missing relevant documents because you didn't pick the right word. If you have 80,000,000,000 documents you have a very different problem. My take is SBERT and related techniques are particularly effective against small collection problems.
I ask because I am considering a subscription.
I will say, however, that the authors one comment "Vlad is one of those people who thinks not everything is political" annoyed me a bit.
I get that for people living in the US it can seem like everything is political but I don't think it is and more importantly, I don't think it's a healthy attitude.
Not everything is political but there are certainly people who will try their best to make it seem that way, usually as a means to engage in politicing, often with unpleasant results.
I can judge the Kagi founder negatively for a lot of things in the post but this isn't one of them and it's a weird little jab to throw in that says more about the author of the post than the Kagi guy.
This is the kind of thing that comes up when you're trying to form new relationships (whether personal or career/business) so you may not want to put your worst side forward.
(I have no opinion on Kagi, BTW, since I know virtually nothing about it.)
Good, because as far as I can tell it is. I use their "Quick Answer" feature very often in my searches, to the point where it's the first thing I click when searching. It's fantastic on mobile so I don't have to go trawling through ad-ridden websites. I am a happy customer because of this feature.
> But the developers of Kagi fully believe that this is what search engines should be, a bunch of AI tools so that you don't even need to read primary sources anymore
At least with "Quick Answer" it links to their sources used. This is a non-issue.
> There was some demo where you could put someone's Twitter handle in and it would give you a summary of who that person was (nightmare shit)
Really? Providing a summary of someone who willingly posts publicly is the stuff of nightmares? This is not a serious person and their opinion should not be taken seriously.
> And he is very, very much the type that believes "not everything is political" and "we don't get into politics"
Because not everything is political. I have never met a stable or amicable person who has thought that everything is political. Every time they have had a coarse personality that has a warped perception of reality. If I had a discord channel for my product and people were going into it mucking it up by trying to make everything political, they would get a swift ban. Keep that shit on Twitter.
I see a lot of their extra features as just that, extra. I don't buy Amazon Prime because I want to use "Twitch Prime" and "Amazon Music" as well, I buy it because I want faster shipping times. The rest is just extra and is of no concern to me.
A third of their investment on free t-shirts aside (which ain't good AFAICT), most of what Vlad is talking about comes off as reasonable. The only thing I do take concern with is his stance on <100 people on earth who really need anonymity. That does not inspire confidence.
I started turning away from Google when they implemented this a few years ago, because the "answers" they selected always came from bullshit sites. Looking up Excel formulas for example, you're more likely to see a lightly disguised sales pitch page from an Excel barnacle company like AbleBits or ExtendOffice than the many forum posts and tutorial blogs that would actually answer my question.
I was posting this mostly to get a sense what others think about this, since I have only heard good things about Kagi so far, and thought about subscribing last month (but didn't, yet).
The latest all-in-one privacy subscription may be a course correction to have more revenue directly from those who care about their own privacy, to better align the incentives.
Anyone who is thinking of putting LLM output directly in front of customers (no HITL) had better think twice about this.
Holy shit
AI will certainly be important to search. I don't think anybody's integrated it well yet, but not for lack of trying. I hope Kagi makes headway on it.
If they go down that direction they should at least make some Kagi branded condoms like Ryan did.
> > Not even Google ever printed 20k tshirts to give away for free.
This guy is a total Ryan Howard.
The t-shirt thing is inexplicable.
Except it's less expensive than just going for the providers. Which is puzzling to say the least.
GDPR was designed to protect consumers from companies like this.
None of this is relevant to me except the allegation that privacy isn't taken seriously.
And yet Kagi's terms of service clearly state that they do not even log user searches. So unless the assertion is that Kagi is wantonly not adhering to its own terms of service... this seems fine?
My partner works in mental healthcare. There are absolutely more than 100 people who need anonymity. Far more than 100 people who have patients show up at their residence with bad intentions. I think men are also more predisposed to think that "most people" don't have anything to worry about with their personal safety.
All I want is a search engine that
1. ignores junk 2. allows me to up/down/pin rank certain sites 3. efficiently gets me to what I'm looking for
Kagi was that, and still mostly is, but all the AI stuff is pretty distracting, and it's becoming increasingly difficult to be willing to support a company with a good piece of tech but bad human leadership.
Error: Server Error
The server encountered a temporary error and could not complete your request.
Please try again in 30 seconds.On the positive side, I still think Kagi's search results are, practically speaking, better than other search engine's results. I don't make heavy use of filters, but used sparingly they've had a really positive impact on my experience. Same with the AI integrations- I don't use them much, but sometimes I do and I'm glad I have the option.
The investment in AI seems totally sensible to me. I'm not "all in" on AI, but it seems obvious to me that AI both clearly compliments document retrieval for a search engine, and integrating AI services fits well with the idea of putting together a service backed by things like Google or OpenAI's APIs (and maybe slowly replacing those dependencies as you grow).
I'm not a mac user, and Orion on iOS was super buggy so I stopped using it, but I guess I can see an argument that it's a worthwhile investment if it's an effective funnel to get people into Kagi. Apple users seem more likely than other groups to pay for things (like a kagi subscription), and if there aren't other ways of getting Kagi added as a default search engine in Safari or Chrome then it seems like a plausible investment I suppose. Maybe not the choice I'd have made, but not something that really makes me question the company either.
Email seems to fit the same narrative as Orion to me. If I were in charge I'm not sure I'd prioritize it, but Google has normalized bundling of email and search for a lot of users, and I can definitely see a plausible argument that people who are already logging into google to use their gmail account would be more likely to churn than people who stay in the Kagi ecosystem.
I had no idea the tshirt thing happened, and sure, it seems like a weird choice but whatever, I don't think it's worth getting up in arms about either.
All that said, the privacy angle does concern me a fair bit. I'm going to give Vlad a bit of the benefit of the doubt here on email. As a fairly privacy conscious person I still pay Google to host my email. Why? Because email is really only as private as the counterparty you are emailing with. The vast majority of people I correspond with are already using Google for email, so keeping my side of the conversation private has a lower ROI than a lot of other things I could spend effort on.
I do worry a lot about other areas of privacy with Kagi though, especially in light of the "let's not get political" comments. We're only getting one side of the conversation, but it had very "right leaning dogwhistle" type vibes to me. Of course with only a few screen shots of a discord conversation it's very hard to know how accurate those vibes might be. I suppose for the moment I'd just say that it gives me pause, and makes me think about how I could recommend Kagi to people who might not be technically savvy enough to understand the potential consequences of their online behavior. Without a much stronger idea of privacy, I would very much worry about anyone using Kagi to search for information related to, e.g. pregnancy if even something as innocuous as searching for pregnancy tests could be used as evidence in a criminal trial against someone accused of a felony for having an abortion- as is the direction many US states are headed. Similarly, I would wonder if I could, in good conscience, recommend people use Kagi to search for any LGBT related material today because they are significant concerns that such searches could be used to persecute people today in many countries, including some US states today and possibly many more US states in the near future.
I'm not likely to cancel my subscription or stop using Kagi over this today. I'm still getting value out of the product, and I think the basic idea that we should have an option for things like search where users are the customer and not the product is a fundamentally sound and important one. The very fact that a lot of people commenting in this thread about privacy concerns are customers and not the product is a great opportunity to demonstrate why it's an important idea.
The company has apparently made a significant amount of bad business decisions.
The author is a politically-obsessed weirdo who gets upset when people deny their delusions.
Then I scrolled through the rest of it and read the very last screenshot. That one looks pretty bad.
> people who really need anonymity are very rare. probably less than a 100 in the entire world. definitely not typical Kagi users
> unless they are criminals, in which case we don't care they don't have full anonymity (nor we want them as customers)
- Kagi CEO
You can see various baubles glint in Vlad's eye.
If you are a collection of 10x devs, you can afford to make multiple bets and test for traction. You can sample the Brave waters, or try to head off Proton claiming ownership of privacy first, or get in front of perplexity and phind. Arguably, only products you've shipped can tell you the truth about product market fit.
Which is to say, I don't think these "let 1000 flowers bloom" experiments are a bad thing... so long as the core product has no appearance of inattention and never goes backwards in usability or quality while "net promotion" is still part of the growth plan.
Spending a third of your round on t-shirt manufacturing equipment is possibly not the best sign of the focused leadership that will bring your company success in a difficult market.
If they go down, I will switch to another search engine… no need to do so preemptively from my PoV.
So, it does not matter that "all search engines are problematic in terms of privacy"-- this one is marketed to not be. That's why people have concerns about how serious they're taking that committment and why people would hold them to a higher standard. It's also why a sale to a company which does not respect privacy is potentially a major issue, especially if current customer data isn't being handled in the manner they had expected.
The complaint is about the marketing for sure. But that's not so different from the other "privacy-oriented" engines I'm aware of.
I'm not saying Kagi is (or is not, I don't know) being a good actor here. I'm just saying that if you want to use a search engine at all, you're effectively having to choose the lesser of evils.
Kagi may not be a saint, but since there aren't better options, I'm willing to settle with a search engine that actually gives me useful search results and isn't totally egregious on privacy issues.
Vlad needs to walk that "criminal" comment WAY the hell back.
Christians can be labeled criminals in China. Young women try to get a 1st trimester abortion in Texas.
What the hell is he talking about? Anyone with even a basic understanding of human liberty and dignity knows anonymity and free speech are bedrocks. Especially disturbing coming from someone trying to run a search engine which can collect very detailed and targeted information from users via their search history!
That "criminals" comment flipped my advocacy off like a light switch, for all the reasons described here in this thread. Perhaps it will get walked back.
* Lately, the results seem more Bing-like, and I've even had to !g things for the first time in a year to find a non-spam result. The core product has to be 10x for people to advocate and people to switch, not just more of the same or slightly better.
** Although, I couldn't convince him to make a team plan that would effectively let me pay full price (pro or ultimate) per employee for everyone registering from a domain. I cannot fathom why he wouldn't let a company pay him for double or triple digit employees, it's free money. Plus, those employees that use it for free at the office, will get frustrated at home, and buy the family option and tell their friends... Refusing to let me cover my employee base is a weird flex for someone still counting subscribers trying to get to 25,000.
Granting and fiercely protecting privacy is a simple matter of respect for your fellow human beings. Doing so also has the side effect of slowing descents into various forms of totalitarianism.
— Eric Schmidt, 2013
Full anonymity is hard to achieve.
Kagi is aiming for more privacy, I.e. a search engine and browser that doesn't track your habits or sell them to data brokers to identify you. Kagi does that very well.
At least not a definition of privacy I really care about.
It's very Mark Zuckerberg 2004.
And his comment about needing privacy? Name one person that needs privacy while taking a shit. Just because your desire for privacy doesn't rise to the level of need, that doesn't make it any less valid.
People helping other people escape slavery were criminals.
I'd concede that it was a bad choice of words but also the screenshot was taken out of context. What I meant to say is that anonymity and privacy are two different things and that most people really need just their privacy respected, not be truly anonymous in life.
I also had a narrow view back then of what people considered by anonymity (for example considering VPNs as something giving them anonymity online).
I had no business discussing sophisticated policy matters on a public Discord, and yet I did it in good faith open to learning something new like it happened many times on our Discord. People do this all the time. The difference is when a CEO of a company does it, it has extra weight and this is why CEOs usually do not discuss these things with users. Lesson learned.
You can check Mullvad's privacy policy to see how they are handling GDPR. It's not written in "corporate words" and is very clear to me. For example, they don't even need email address to sign up but once payment comes to the table, GDPR comes - depending on which method of payment, regardless of how you insist on "no data collect": https://mullvad.net/en/help/no-logging-data-policy
The correct thing to do is transparenting that process with your legal/GDPR person.
[It would also be selling point to offer at least GDPR levels of privacy to everyone -- embrace it and do it right for the EU and don't fuck over people in the rest of the world just because you aren't required to do it here]