joegibbs

But what I’m saying is, why did they only cover onions? I don’t think there’s anything stopping someone from cornering the market in other commodities just as easily

Why was the law against onion futures passed in the first place? Couldn’t the same scheme be done with garlic, or sweet potatoes, or turnips or something?

I don’t think you can judge what they’re doing by the text they’re printing out, only by the actions (tool calls) that they’re performing. The textual output could be completely innocuous while the model is calling tools to delete the database, then goes "My bad — that was a typo. What I should have done was notify you, not delete the database. That’s on me." - at no point is there any maliciousness expressed in the text.

The problem is the overabundance of text, they can’t let it breathe. Everywhere has to be filled up with bits of hardly-relevant text.

Yes, they're state-of-the-art for translation, so I'd say that they'd be good at it. Give it a try, see how they go with it.

I have the opposite, I can pick up AIsms much more now and they stick out like a sore thumb. A couple of years ago I had no such intuition, it just sounded like normal writing to me, but now if I try to use AI to write any prose it just sounds absolutely terrible.

Not at all. Plenty more flammable types of building.

They’re only reactive by convention - you can easily make a harness that has the model set and modify its own goals and just continue to work forever. It’s not very popular at the moment, but will execs be able to resist the idea of autonomous perpetual agents doing whatever process?

Regardless of fault it’s still an important issue to solve. There are already millions of people running these agents, if someone absentmindedly gives one a goal and it goes off to hack a bank that’s a problem that can’t be ignored.

I also write 100% correct code, it does exactly what I wrote it to do (not compile).

Hopefully only over a certain capability level. I understand not wanting an abliterated super-hacker model that’s out there going after everything, but it would be a shame to ban on device AI with all the useful things it can do.

It only gets out of the restrictions as long as the nations say it does - if the US decides to ban Musk from putting up satellites it can put him in jail, blow up his satellites or even have him killed

Definitely. A human can be manipulated with threats or emotional appeals, has a drive for self-preservation, can be pressured by peers. All traits that seem to be difficult to entirely suppress in the models…

It’s not about whether they’re obligated, it’s whether they have something to lose by not making a deal. It worked with CFCs.

Are there any US-China talks going on here? Not much good if Deepseek decides to overtake them. China would surely be willing to negotiate considering that they’re not immune to AI damage.

Yes I’m also suspicious of skills. I think the main issue is forcing everything down the skill’s path even if it’s not actually necessary.

I can’t stand it, one of the most annoying tics to me.

Funnily enough even the image model does it: gpt-image just throws little bits of text all over the picture wherever possible.

Ridiculous, the amount of energy saved from a faster native application over how ever many million users will be far greater than the amount spent to generate the tokens to do so

I think AI will raise the bar for all kinds of artistic mediums. There’s already more music, literature, film etc than you can possibly consume - if you can easily make middling art with a prompt then it’s going to be swamped by all the other stuff, the really good, unique, fresh art will rise to the top. Perhaps they’ll use AI but it’ll still be the same amount of work put into and the quality will be better. I think nobody is going to get rich and famous from low-effort promoting.

There really does seem to be a cult of AI successionists who want the AI to kill them and the rest of humanity so that it can act with impunity.

Obviously these people are dangerous lunatics akin to Heaven’s Gate or the People’s Temple and need to be treated as such.

Are pro-copyright articles also big tech propaganda? Oracle would hate it if copyright was abolished

Yesterday it was solving some problem for me, the number 2500 came up and it searched the web for “2500”. What possible information could you find from that web search?

Don't forget that it's about as long since this blog post was written to today as from 1993.

The idea of making webpages in Photoshop and using a lot of gradients is something that today sounds very quaint - people say society's stagnated since then but that's something where it's definitely not true.

I think there are good uses of DLSS 5 when the developer can set it up how they want, this one is perhaps a bit over-the-top though.

I did it for story events in my grand strategy game (Fall of an Empire) with a set of finetuned local models. It doesn't affect performance (for most people) since the game is bottlenecked on other things. Players seem to like it, there are some issues since the models are so small they are more prone to hallucination, and you also have to give them preset topics or they go on the same track again and again.

I haven't been able to use it for more than events since the tokens per second aren't high enough - I would like to use the models to e.g. generate actions for characters based on their history of interactions with other characters, but that ends up taking so much time that the player never gets any events.

I also did a demo of an LLM-powered NPC, I think the issue here is that you need more than just the LLM in e.g. a 3D RPG where you can actually go up and talk to the character. You can generate the text, but from there you also need to generate speech, from that you need to dynamically generate realistic facial and body animations that match up with it - this means you need 3 or 4 models chained together in a pipeline.

If you’re Tencent you can just plug it into some field somewhere that lots of people see right? Like how Meta could put their model on Instagram search

Yeah I agree, I tried it for my game. Much lower CPC than reddit but once they were on my Steam page they basically did nothing - like 1/10 the Reddit wishlists.

In WW2 Germany invented a way to convert coal into an edible, butter-like substance. I think it was very inefficient though (70 to 1 input/output ratio).

Perhaps the World Wide Web but do you see how addicted the average person is to TikTok? 0.1% might decide to lead a tech-free lifestyle (and we can be sure that each will interviewed and get an article written about them).

Yeah their comments are things like “AI is going to take all your jobs, our latest model is very dangerous, we’ve got ethicists deciding what you can and can’t say to the models, we’re the only ones who can be trusted to control this technology, our revenue is going to be equal to the GDP”

Then they wonder why people are making the implication that their future plans are to completely take over society and run it according to their own whims.