Regarding the bookmarks bar, Settings / Appearance / Show Bookmarks Bar. If the setting is off, the bar only appears on new tabs. I found that by accident.
261 karma · joined May 29, 2014
There's a long history of that sort of behaviour. ISPs gaming bandwidth tests when they detect one is being run. Software recognizing being run in a VM or on a particular configuration. I don't think it's a stretch to assume some of the money at OpenAI and others has gone into spotting likely benchmark queries and throwing on a little more compute or tagging them for future training.
I would be outright shocked if most of these benchmarks are even attempting serious countermeasures.
I echo the parent comment. I'm really on a Proton user for email and VPN. The quality drops off rather quickly after that. Calendar, Drive, Pass, and Wallet are all adequate at best; their primary selling point is not being Google rather than being particularly well built or supported. I would rather see them focus on being a truly competitive ecosystem.
I'm also not terribly impressed at the way they've positioned Lumo as a separate service from the existing Scribe AI features, and so conveniently not part of Ultimate plans.
Google's approach works well enough when you're searching for a commodity and you don't care terribly much about the specific source. I get the impression Google, especially post-LLM, wants to divorce satisfying your question from the underlying sources.
I find Kagi is better at finding a specific thing, especially if you're willing to engage with it as a tool, ye olde search engine style. If my query doesn't find what I want, it's usually apparent why and I can reframe it.
Google's case (and Meta and spam calls and others) is at least in part an incentives problem. Google hasn't been about delivering excellent search to users for a very long time. They're an ad company and their search engine is a tool to better deliver ads. Once they had an effective monopoly, they just had to stay good enough not to lose it.
I've been using Kagi for a few years now and while SEO spam and AI garbage is still an issue, it is far less of one than with Google or Bing. My conclusion is these problems are at least somewhat addressable if doing so is what gets the business paid.
But I think a real long term solution will have to involved a federated trust model. It won't be viable to index everything dumped on the web; there will need to be a component prioritizing trust in the author or publisher. If that follows the same patterns as email (ex: owned by Google and Microsoft), then we're really screwed.
But compared to using Kagi, I've found found LLMs end up wasting more of my time by returning a superficial survey with frequent oversights and mistakes. At the final tally I've still found it faster to just do it myself.
I will say I do love LLMs for getting a better idea of what to search for, and for picking details out of larger blocks.
I think you're putting too much emphasis on The Algorithm. It's a problem, and I agree it's probably the worst offender, but similar problems were observed decades ago with children (and adults...) allowed to watch too many hours of uninterrupted TV. Cutting back to chronological feeds might improve some things but I don't think that's the root of the issue.
I would suggest the primary difference between then and now is accessibility. As a kid, my screen time was limited not just by my parents indulgence but the social pressure from using a shared device. Smart phones let you carry your personal distraction with you.
I agree they are a wonderful invention but I'm not sure grade school students need to be connecting to anyone, anywhere throughout the entire school day.
The talk about fentanyl, the supposed trade imbalance, and all the rest is just smoke and noise, because if he said that he wanted to raid Canada for resources and Lebensraum it might actually threaten his support.
He's not being subtle in this. He and his people have made constant comments that Canada doesn't work as a nation, that the border is artificial, and even explicitly that the tariffs would just go away if Canada became the 51st state. This is economic warfare driven by manifest destiny.
That's also each American citizen getting... $23.27 each year for 10 years.
People are just awful at conceptualizing large numbers. Assuming that number is accurate, $7.9 billion is wasted in the process of delivering $1.6 trillion in 2024. Waste and loss scales with the work being done, it doesn't care that you got sticker shock.
I challenge you to find anything you do in your day with greater than 99.5% efficiency. You waste a higher percentage of the food you eat, stuck to the pan.
I'm very much in favour of removing the guardrails but I understand why they're in place. The problem is attribution. You can teach yourself how to engage in all manner of dark deeds with a library or wikipedia or a search engine and some time, but any resulting public outcry is usually diffuse or targeted at the sources rather than the service. When Claude or GPT or Stable Diffusion are used to generate something judged offensive, the outcry becomes an existential threat to the provider.
The implication isn't that drivers get steadily better as they age but that the longer your history of safe driving, the less likely you are to become a liability.
Historically radio worked because their was actually money in the advertising, but also because radio was inherently regional. You could license content at a rate that made sense for your station's reach.
Now it's trivial to broadcast over the web to a huge audience but the floor has fallen out on monetization. Podcasts and streaming both work because they don't rely on licensed content, but music can't make the jump.
Is corporate misinformation so much better? Recall about Tienanmen Square might be more honest but if LLMs had been available over the past 50 years, I would expect many popular models would have cheerfully told us company towns are a great place to live, cigarettes are healthy, industrial pollution has no impact on your health, and anthropogenic climate change isn't real.
Especially after the recent behaviour of Meta, Twitter, and Amazon in open support of Trump and Republican interests, I'll be shocked if we don't start seeing that reflected in their LLMs over the next few years.
This is the classic "solution looking for a problem" you see so often with NFTs and blockchain generally. The proposed problems (in this case verifiable and transferable ownership) already have mature, satisfactory solutions.
NFTs don't benefit consumers (because the tangible value exists in a centralized server somewhere anyway) and they don't benefit publishers (who can already implement those features if they cared to). They benefit people that want to get rich speculating on NFTs.
Reddit is a cluster of independently operated communities that occasionally get signal blasted onto the main feed. The top moderator of a subreddit owns it and sets any standard they like.
With Bluesky there are no defined communities outside of collective engagement with particular topics or hashtags or networks of follows. There is also a "Show me [more/less] of this" button every post, and so far it seems like the platform is pretty solid at respecting your preferences. They also seem to be actively moderating the really open bad behaviour off the platform.
I don't think Bluesky is as good as a well-moderated subreddit for long form discussion, but if you spend a little time curating your feed I think you might enjoy it.
American Exceptionalism at work...
Miraculously, carriers simply started offering "tabs" or other language where you pay the subsidized phone cost as an addition to your plan bill for the contract period, with a clause that if you cancel early you still have to pay up the difference.
Arguments in favour of locking are nothing but corporate apologia and business crying wolf.
KeePass was a great bit of software but managing the vault syncing myself and having to wait for (and trust) the third-party Firefox extension to update was tiresome. For about a buck a month, LP was a pretty good deal and handled all of that overhead for me.
I eventually moved to 1Password and it's still what I recommend to most people. $45CAD a year is a pittance for how often I use it. The app and extensions are always up to date, they "just work" even for my 70 year old father. At $12CAD a year, Bitwarden is pretty damn reasonable too.
I don't get the hand-wringing when it comes to reasonably priced services. Development and infrastructure costs money. Yes, a power user can manage everything entirely with free software and a portable sqlite db but that isn't sensible approach for the vast majority of people.
I remember playing a ton of awful shareware games as a kid, on floppies and disks packed with a few dozen each. Same thing later on, bringing a game home from the store only to realize I'd chosen poorly.
And it's easy to get a bit jaded after years of experience and look back with a rose tint. I loved the hell out of that junk as a kid, because it was still the best stuff I'd yet seen and I had time to spare. Today I'm a little pickier.
The pricing seems to implicitly acknowledge this: $35/m billed monthly vs $8/m billed annually! Would you really expect anyone to intentionally renew monthly? I can't argue that people forgetting to unsubscribe pays the bills, but as a business model it leaves a bad taste.
The UX is typically better, and that was especially true back when Gmail was taking over the market and introduced everyone to the concept of never needing to delete old mail. Good looking apps are available for every device, with minimal setup. Deliverability is so good most people have no concept of ending up in their contact's spam folder. It's what your friends and family are using, and its free.
Concerns like privacy or being able to change provider without losing your address are abstract and out of sight.
Either you insist that copyright must be respected at every level, and the creators of material used for training deserve appropriate compensation, or
You throw out copyright completely in this context, but that means the resulting models cannot be treated as proprietary either unless they were produced using absolutely no unlicensed training data.
I think there is an argument for both. Want to create a proprietary model for commercial use? Pay up. Creating an open source, copyleft project exclusively for personal use and artistic expression? Exemption.
The current status quo is perfectly described by powerful corpos extracting rent. Billions for themselves and pennies for the average artist.
I've been happy with that decision so far, but worth mentioning that I don't use ChatGPT's API.
I'd recommend anyone suffering play around with the generators on myNoise. It wouldn't be an exaggeration to say they kept me going when I was first adjusting to the ringing.