wrasee

I'm confused, partly out of my own ignorance but after looking at their site I'm still none the wiser. Do modern cars essentially have some interface that means you can control them electronically? Is this a public interface, like an official standard shared between brands or are they adapting per-brand? Or perhaps even revere-engineering the car's internals in some hacky way? What about from the car manufactures, do they support such things? Would installing this affect your Car's warranty? Are you "supposed to do this"?

And what's their safety record? I guess if they're not saying, it's not something they want to talk about. Or worse, they don't even know because I doubt users are self-reporting.

So many questions. I spent some time on their site but got frustrated and bounced before i could get any real information. It's just talks 'car compatibility' and 'buy now'. Seems to present as a magic box you should just trust.

And while I'm sure I could find answers the fact that these things aren't front-and-center, at least to me, doesn't send the right message. I don't want to just "plug in and go" with something so safety critical. Seems to market as more toy/gadget than "safety first".

Same with Tesla, mind.

It is a remarkable achievement. But if “some training data from OpenAI” turns out to essentially be a wholesale distillation of their entire model (along with Llama etc) I do think that somewhat dampens the spirit of it.

We don’t know that of course. OpenAI claim to have some evidence and I guess we’ll just have to wait and see how this plays out.

There’s also a substantial difference between training of the entire internet and one that very specifically targets your competitor's products (or any specific work directly).

Right. I think many here are enjoying the Schadenfreude against OpenAI, but that hardly makes it right. It just makes it a race to the bottom.

Yes, but in science you reference your work and credit those who came before you.

Edit: I am not defending OpenAI and we are all enjoying the irony here. But it puts into perspective some of the wilder claims circulating that DeekSeek was able to somehow complete with OpenAI for only $5M, as if on a level playing field.

If someone replies to your comment then I think it’s entirely fair that they take your point in the context in which it was intended. Otherwise, if they are not interested in the point then simply don’t reply to it.

No-one enjoys being taken out context.

But I do accept that given the hostility of replies I didn’t make my point very effectively. In a nutshell, the original comment was that it’s surprising a small team like DeepSeek can compete with OpenAI. Another reply was more succinct than mine: that it’s not surprising since following is a lot easier than doing SOTA work. I’ll add that this is especially true in a field where so much research is being shared.

That doesn’t in itself mean DeepSeek aren’t a very capable bunch since I agree with a better reply that fast following is still hard. But I think most simply took at it as an attack on DeepSeek (and yes, the comment was not very favourable to them and my bias towards original research was evident).

Anthropic was founded in part from OpenAI alumni, so to some extent it’s true for them too. And it’s still taken them over 3 years to get to this point.

Of course not. But in this context the point was simply that it’s not exactly a fair comparison.

I’m reminded how hard it is to reply to a comment and assume that people will still interpret that in the same context as the existing discussion. Never mind.

I would extend the same reasoning to Mistral as DeekSeek as to where they sit on the innovation pipeline. That doesn’t have to be a bad thing (when done fairly), only to remain mindful that it’s not a fair comparison (to go back to the original point).

Disrespecting copyright and personal data is good for users? I guess I disagree. I would say that it’s likely great for the company’s users, but not so great for everyone else (and ultimately, humankind).

Mistral.

I don’t think it’s that harsh. And I don’t also deny that they’re a capable competitor and will surely mix in their own innovations.

But would they be where they are if they were not able to borrow heavily from what has come before?

This is true, and practically speaking it is how it is. My point was just not to pretend that it’s a fair comparison.

Also don’t forget that if you think some of the big names are playing fast and loose with copyright / personal data then DeepSeek is able to operate in a regulatory environment that has even less regard for such things, especially so for foreign copyright.

Except it’s not really a fair comparison, since DeepSeek is able to take advantage of a lot of the research pioneered by those companies with infinite budgets who have been researching this stuff in some cases for decades now.

The key insight is that those building foundational models and original research are always first, and then models like DeepSeek always appear 6 to 12 months later. This latest move towards reasoning models is a perfect example.

Or perhaps DeepSeek is also doing all their own original research and it’s just coincidence they end up with something similar yet always a little bit behind.

Agree with all of that. I would love for Kagi to be the future of search. Let’s hope that they can do so at $5/month.

My hunch is that they are happy just growing a sustainable business. And I say good for them.

I’m not. I’m pushing back against folk that forget not everyone thinks the same as them. That “$5 is practically nothing” reads like a statement of fact, which is obviously ridiculous depending on where you’re from.

And yes, it definitely has a “let them eat cake” vibe which is absolutely worth a push back.

A strange thing to say about a discussion forum, whose essential purpose is the discussing of things. There is absolutely social value in understanding what people are willing to pay for things. And if it’s cost me more than $10 to present my view, well let’s just call that a donation. You’re welcome.

And people spend a tremendous amount of time discussing the price of essential goods like groceries. Economists go bananas of it. Especially in some regions with a cost of living crisis. What a strange example to make.

Except for these types of services scale is in your favour. A bakery has a high cost of goods sold, whereas services like this tend to have high fixed costs but relatively light costs per customer. So scale is very much to your advantage (within reason).

I suppose my argument was essentially that $5/month felt on the high side, so I love that the answer is actually to pay $10/month. This is the way.

In all seriousness, I am sure you are right and don’t deny the promise on offer. I’d also be willing to pay for search much more so than many other online services of more questionable value.

But I just don’t think they’re going to get cut-through at $5, not least $10, in the way they might exponentially do so if they could stretch to an entry tier that really starts to feel accessible to a much, much wider set of people.

WhatsApp took over the world at $1/year.

If they get 100x the customers on a $2 tier than a $5 tier then yes, they can be profitable.

The question is, could they get 100x? Or even 10x? I don’t know, and I assume they’ve done their homework and believe that they can’t. I just wonder, what if they’re wrong?

> $5 is practically nothing

That’s quite a privileged thing to say. This forum is read by folk all across the world and I would say that it’s up to each individual to say how much $5 means to them.

Genuinely interested. So I went to their site and I have to say it's quite refreshing.

But then - a few bucks a month turns out to be $5/month, and that comes with a limit of 300 searches/month. That's an average of just 10 a day and while in practice I probably wouldn't actually hit that limit it's uncomfortably low enough that the added fiction of each search having a tangible cost is just a non-starter.

I can't help feel that when the competition is free - and on precisely the recognition that people don't pay for search - that a more aggressive $2 tier for personal customers could get a much wider penumbra of folk such as myself over the line, rather than be yet another $5 service that when applied consistently actually start to add up.

I'm a big linux fan. Use it every day. I even use linux on the desktop to develop Windows software that isn't even distributed on linux.

But am I really that alone in recognizing a lot of truth in this article? It's sad to see so much hostility within some of the other comments, even here on HN, quickly coming to linux's defence rather than having at lest some degree of reflection and acceptance towards what are almost undeniably the very common experiences of many linux users.

It somewhat proves point 7 that while not exactly to the letter of the examples given, support the point that until there is wider acceptance of the problem we are far far from a solution.

It's possible to think very positively of linux yet still recognize so many of the challenges that make linux hard to use on the desktop. This very day I can't suspend/resume my desktop because the kernel crashes on resume, a pretty basic feature of a computer in 2024. I put up with it, because Linux is still a great choice for a technically-minded user. But there are other software orgs out there that simply wouldn't tolerate this level of lived experience of any end user. But linux does, and we are where we are.

This is so on-brand for Microsoft.

I can't help but believe that their aggressiveness in pushing AI to everyone is going to hurt them long term, especially in terms of trust. We'll see.

I am saying that it's unhelpful to miss the context, and it would be a shame if one was to misread that the Swift compiler can't even provide error locations, which it obviously can in the vast majority of cases.

Swift's compile time performance is obviously well documented. It's not great, and performance is definitely one of the trade-offs. SwiftUI leans heavily on the compiler and is very much at the sharp end of this. Most "normal Swift" code is by and large just fine. Of course discussions tend to spiral outwards, but stripping context and adding largely uninformed, quick judgements rarely adds value to any discussion.

You’re aware that the comment was specific to (I would say) reasonably complex SwiftUI expressions? How familiar are you with the context here?

C++ is hardly niche, nor I would say is the desire to interop with large, established languages with decades of existing code and libraries.

As someone who has worked in both Swift and C++ I’m grateful for those that choose to make an effort in this space. And of course, just because some do doesn’t take away from others that work on SwiftUI. Both can be true at the same time.

But even then the weighting is _very_ uneven. The number of votes per elector can vary wildly by state, by as much as some small whole multiple. So the “weight” of one vote in one state can be say, four times that of another state.

It’s amazing to me that this can stand and efforts to change never seem to get very far.

Ah yes I remember that now, I had forgotten about that!

Funny, especially now that I see Apple are now going the other way with a dedicated "Passwords" app on iOS 18 and macOS 15. And for Apple to do this - against their instinct for featureless simplicity and implicit integration - to give passwords their own "shop front" as a dedicated app I think really does acknowledge the first-class importance that passwords now have, even for a broad audience.

It's a shame as I think Mozilla could really compete well in this space. They are both cross-platform, have their their own browser and have a good reputation on privacy. It's a killer combo. Bitwarden is evidence you can make it work and you don't need massive big-tech budgets to make a difference.

If Mozilla released a separate passwords app so you could manage and access your passwords outside of Firefox I think the two would be more comparable. That would promote your passwords as part of your Mozilla account, not just Firefox.

Bitwarden excels here, and i think is the model to beat. However, Mozilla would have the advantage since their browser integration would essentially be built-in and first class.

Otherwise, unless you use Firefox exclusively for everything I just don't think a single browser is the right place to manage passwords. I would say that's true even for a broad audience, given the importance of passwords and security in the modern age.

Bitwarden is also nice in that you can "lock" access to your passwords while keeping the browser open. That way, for the 99% of the time you're just browsing the internet you essentially don't have access to all your passwords "open". The last time I looked at this I had to enter my master password on opening Firefox, even if I didn't need access to my passwords. That meant that "unlocking your vault" is essentially tied to opening the browser. That alone was enough for me to bail on it.