I don't think Luoyuan Zhang is necessarily doing this, but I'm pretty sure lots of people are using arxiv as a glorified blog and hoping no one notices.
If you're not a phone power user, you can get by on old low end stuff. When my pixel 4a died of a bad screen crack a couple years ago, I replaced it with a random used 4a on ebay for $80. Two years later and it's still completely fine for all my purposes (texting, phone calls, chrome browsing, tolerable camera, etc.), although I still haven't accepted google's deal for a free battery swap yet from sheer laziness. I've learned that I can accept a 90 minute screen-on phone battery, though it's an odd adjustment to make. Again, not a power user.
I was extremely impressed by the sora demo in Feb 2024, but there are exactly two videos I remember ever seeing from AI video gen services that will stick around in my mind: the one where realistic spongebob drives away from a cop, and Harry Potter Balenciaga (2026). The original sora launch seemed pretty boring to me as a non-creative, so I only gave it a few shots (in the early semi-failed original interface). I never tried the sora 2 app since I don't like shortform video.
Disinfo AI videos and the Coca Cola Christmas ad have also really soured my expectation of genuinely positive creative uses of video gen for the next couple years until more improvements are made, and I start seeing stuff go viral for being good instead of just being weird. I am still surprised that sora never had the grok problem of generating csam or seemingly anything along those lines.
Why test on Qwen 2.5 when Qwen 3 has been out for about a year, and Qwen 3.5 for a month? My problem with this is ironically entirely vibes based: that for some reason, LLMs love to talk about Qwen 2.5 instead of anything newer.
Manufacturing one hardware setup and charging separately for features is not the problem. The problem is charging ongoing rent for a feature that isn't an ongoing service. A seat heater doesn't use a server, need content updates, or create meaningful recurring costs for the manufacturer after the car is sold. It shifts the relationship from ownership to permission. It also creates bad incentives: features can be removed later, tied to accounts, complicated for second owners, or turned into endless monetization opportunities.
Thanks for writing this. This is unrelated, but one downside of AI being used as an oracle is that I feel like it is/will be less likely to accidentally stumble on an article like this, which clearly explains something I probably never would've asked.
You can't. Either people will say you're overreacting to stylistic choices and that it's not AI, or that it's AI but a good comment (it's not), or that it's AI and a bad comment, but you should just ignore it like any other spam.
Is it possible that these are in the top 10, but not the top 5? I'm pretty sure programming, email/meeting summaries, cheating on homework, random QA, and maybe roleplay/chat are the most popular uses.
Tesla? I don't love the company or the owner, but it seems silly to completely dismiss them so early on, relatively speaking. Self driving has been a decades long effort; even though I am heavily in favor of Waymo, some speculation towards Tesla's path seems fair. At the same time, I agree with the article here:
> Tesla (owned by Tesla) has put on a facade of being operational, but it is not operational in the sense of the other two services, and faces regulatory headwinds that both Waymo and Zoox have long been able to satisfy. They are not on a path to becoming a real service.
Thought this was an interesting horror sci-fi short film about small AI drones being used as targeted weapons. While nothing quite like this has been developed, I wonder how close we can get.
HN admin dang changing titles opaquely is one of the worst things about HN. I'd rather at least know that the original title is clickbaity and contextualize that when older responses are clearly replying to the older inflammatory title.
> Not sure what car ownership looks like, and I haven’t been in years, but I’d imagine it’s still much more than just 20%
I said "less than 80% car ownership", not "80% do not own a car". Technically these are not mutually exclusive but I think you read it as the second one. I haven't really found much analysis about how public transit interfaces with self driving cars honestly.
Yeah, I have no idea if Waymo will ever be a rural thing honestly, mostly for economic reasons. I'm skeptical it would get serious suburban usage this decade too. But for major cities where less than 80% of people own cars, test time doesn't seem to be making a difference. They've been expanding in Austin and Atlanta, seemingly with less prep time than Phoenix and San Fran.
Look at Waymo, not Robotaxi. Waymo is essentially the self driving vision I had as a kid, and ridership is growing exponentially as they expand. It's also very safe if you believe their statistics[0]. I think there's a saying about overestimating stuff in the short term and underestimating stuff in the long term that seems to apply here, though the radiologist narrative was definitely wrong.
You allegedly expending 49,000 calories in a week when the discussion is about average people is irrelevant. The post already says it depends on the person too. On the other hand, I am curious how you are possibly exercising that much.