jprete

Clearly your training corpus was cut off some time in the 2010s. You're going to need a serious update. Regrettably the Internet stopped caring about text at that time so we're going to have to base64-encode all of TikTok and hope it makes more sense to you than to us.

I know I'm real. For everyone else, I have no way to tell for myself, let alone proving it to someone on a comment thread.

Don't forget text. You can't trust text either.

And no big tech company would run the ads you're suggesting, because they only make money when people use the systems that deliver the untrustworthy content.

It's a message board frequented by extremely tech-involved people. I'd expect the vast majority of people here to have some financial interest in LLMs - big-tech equity, direct employment, AI startup, AI influencer, or whatever.

My point is that the user is adding another layer of abstraction, and that layer of abstraction needs itself to be trusted. When UI elements are really concrete and you can clearly see that you pressed a particular button and the thing you wanted happened, then the UI layer, at least, is a nonissue.

But in retrospect I don't know if my point was that good. The UI problem hasn't actually been solved, and an LLM-based chatbot may actually be more reliable for non-tech users since the user has to do less translation.

I don't have experience with this kind of problem. But I don't think GenAI is the best tool for this, at least not until it's so rock-solid trustworthy that everyone uses such an interface. Even leaving aside AI questions, if I'm looking for a human personal assistant for someone who's blind, and that person will have unlimited access to their electronic life, I'm going to vet that person very, very carefully.

This discussion looks unrelated?

The poetic meaning of the line still works. The sky, the symbol of boundless optimism, the place of open infinite wonder and possibility...reduced to meaninglessness, closed possibilities, and failed promise.

Looking at the database article, I now realize that many of the math and physics articles are written in exactly that style. The authors seem to give complete and thorough maximally-formal definitions first, and only afterwards (if at all) attempt to explain it to people that don't already know the subject.

Thinking about it, this is exactly what I would expect from a process that overemphasizes formal rules of evidence and precision over being of value to the general public.

That's going to be a good standard for a few years, until chatbots are too sophisticated for us to expect average adults to be sufficiently skeptical of their arguments.

You're not wrong, but DVDs were also region-locked.

I think at the time this was both less annoying and more easily circumvented. I think it only came up if you bought DVDs internationally, were moving between regions, or wanted to bring DVDs but no player on such a trip.

DVDs also have less-than-stellar image quality by comparison to today, although a lot of older works need the lower resolution because special effects and the like weren't intended to be scrutinized at 4k.

This is also my strategy. AI content makes me focus even harder on the source of the content instead of the apparent quality, because the current set of GenAI techniques are best at imitating surface-level quality features.

No individual actor - human or corporate - stands to benefit enough because "trust in reality" is neither easily measured nor financialized.

The GP's point matters quite a lot, because the ratio of GDP to non-GDP value is going up. Terms like "financialization" or "commodification" describe a process that makes things visible and measurable. I think this is also what the book Seeing Like A State meant by "legibility".

And the whole reason it's a bad thing is that the human values that can't be converted to a measurable, tradeable commodity are often thrown away, because they're invisible to the commodification process.

You're completely missing the problem. The care assistant, being a fungible and commoditized service provider, can never provide the kinds of illiquid unmeasurable goods that are generated by long-term and durable relationships.

GDP semantics matter enormously for policy decisions because GDP is the easiest metric to use when making those decisions. But it's a metric and highly biased towards measurable values instead of less fungible ones.

The problem of marketisation is that it's a shredder for human values. The economic machine doesn't care what the constituent parts are, as long as they move correctly.

I get the point, but I think the most self-consistent answer is that your conscious experience has a 100% chance of staying in the original body. (And similarly for destroy-and-remake teleporters.)

Broadcast radio stations typically had phone lines for listeners to call. Music stations would take song requests and run contests where the Nth caller would win a prize. Talk radio would put callers on the air to chat with the host.

Maybe you already knew this, but anyone who remembers broadcast radio would also know it. If you're making a joke, I don't quite get the punchline.

I don't mind streaming menus per se because they have about as many extra steps as one would expect given that you need to somehow communicate what you want to watch.

I'm annoyed, though, that streaming app hide "purchased" and rented content in favor of their recommendation lists. If I just rented a movie from Amazon Prime, and immediately open the Amazon Prime app on the streaming box, I don't need recommendations. If anything, I need to be nudged to watch the thing I just bought!

I agree with you that this isn't really specific to capitalism, and is specific to using metrics as proxies for values.

The idea that AGI doom scenarios are really late-stage capitalism is interesting and strikes me as fundamentally correct. The difference between the AGI takeover and capitalism is just the choice of metric to optimize on.

But, I think, it's the act of trying to optimize on a metric itself that is the source of the destruction. Unmeasurable human values can't survive an optimization process focused on measurable ones.

260 hours over two years is 2½ hours per week, or half an hour per working day. Most people have longer commutes than that.

Based on the usage graph, though, it seems like the author had short bursts where they watched a lot of Netflix, followed by long periods of not using it at all. This seems more like they had periods where they just needed to chill for a while?

I also think, since the author is analytically inclined, they should make a histogram of the percentage of total duration of each item they watched. I personally fall into the trap of getting just enough motivation to stream something for 5-30 minutes and then not finishing it; this is a hell of a lot more wasteful than binging a few TV shows.

The actual headline includes an important detail: "The 22-year-old woman who was on the brink of being euthanised – and changed her mind"

Being part of a public court record doesn't seem like something that would invalidate copyright.

I don't see any serious "right way" as you describe it. In particular I don't see a lot of motivation from social groups, and the Internet is horrible for good feedback because lots of people respond to things from a purely emotional place.

Claims that LLMs work like human brains were common at the start of this AI wave. There are still lots of fanboys who defend accusations of rampant copyright infringement with the claim that AI model training should be treated like human brain learning.

I think it's really the AI content that people care about. Devices to disguise a voice have been around for a long time, so I don't think it was a particularly widespread job in the first place.

(And maybe technological changes in that space have always been about style and perception instead of effectiveness? I mean you can always get a paid actor to pretend to be the disguised interviewee and use a radio to relay the answers the show wants.)

The idea that "TikTok is for Gen Z" seems like a very stale meme, although I only have anecdata to back that up.

Microtransactions are way older than 15 years. Wizards of the Coast was selling randomized MtG booster packs in 1993. I'm guessing that the earliest loot boxes for kids were baseball card packs, with very similar psychological purpose to today's game cosmetic collectibles.

I think most PhD programs have the option to leave with a Master's degree after a certain amount of progress. Do you have that option?