Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

AI Built a Nuke and Still Lost

lwilko.com

Post reply on HN
Print view View original

71–80 of 103 posts

Re: AI Built a Nuke and Still Lost

#71

Post by xpct »

titanomachy wrote: post #62
fyredge wrote: post #11

There is something to be said about the qualia of LLM generated passages. Each individual sentence reads as a statement and every next statement a continuation of the previous one. This happened, then this happened... Ad infinitum. Before today, I could not explain to you why AI articles were so obvious to me, but I think I do now. There is no insight to be gleamed. Pre-LLM, authors generally had intention behind the…

It's surprising to me that the big labs haven't fixed this problem. Half the comments here are complaining that this article is egregious unreadable slop, and I agree. Surely with the trillions of investment they could at least figure out how to vary sentence structure a bit and nix obvious tells. Maybe this is just an inherent problem with LLMs.

The following come to mind:

- either the problem is hard, or labs have no incentive to fix it for their main users

- being able to tell that something is LLM-generated, is good

- could be that structure is an emergent property as models get better

Re: AI Built a Nuke and Still Lost

#72

Post by dwroberts »

> I asked the agent what this was actually like for it. It wrote back

Stuff like this just makes the author seem clueless. What is even the function of putting a question like that into an LLM unless you’re already hopelessly in anthropomorphic territory

Re: AI Built a Nuke and Still Lost

#73

Post by roenxi »

fyredge wrote: post #11

There is something to be said about the qualia of LLM generated passages. Each individual sentence reads as a statement and every next statement a continuation of the previous one. This happened, then this happened... Ad infinitum. Before today, I could not explain to you why AI articles were so obvious to me, but I think I do now. There is no insight to be gleamed. Pre-LLM, authors generally had intention behind the…

I think it might be a training artefact of some sort - the current crops of LLMs have never been in a position where they can explore the world as an independent existence and so they might be struggling to model how to explain an interesting experience? The Go AIs had problems with ladders of all things (one of the most basic beginner shapes) back in the early superhuman phases after Alphago. There seems to be some similar and profound gap in the LLM understanding of how to communicate when storytelling.

The "But France was running two clocks at once" paragraph really set me off because I get the feeling something really interesting might be happening that the AI doesn't want to talk about and there is evidence that it is trying to say something. But the result is some amount of gibberish and some amount of vague allusion to something interesting in the prompt context while glossing over all the information that might matter while working hard to create an evocative feeling that isn't interesting. A tense atmosphere with no exploration of why there is tension.

Re: AI Built a Nuke and Still Lost

#74

Post by xpct »

pjc50 wrote: post #20

> I now work with governments around the world at the Tony Blair Institute, which means I spend a lot of time in rooms where people ask the same question: what can we actually trust these systems to do? Oh no - we're going to end up with the Starmerbot 3000. Now I've got the joke out of the way, there's at least four interesting lines of inquiry one could take with this blog post: - teaching the AI how to play Civili…

I very much dislike the idea of teaching the robot to play Civilization and expect those skills to transfer to their advisory nature.

If anything, I'd almost prefer a leader who hasn't played Civilization in their life. Goes without saying that a mature leader could tell these apart, but in this day and age, I'm not so sure whether everyone could.

Re: AI Built a Nuke and Still Lost

#75

Post by artpar »

pjio wrote: post #57
fyredge wrote: post #11

There is something to be said about the qualia of LLM generated passages. Each individual sentence reads as a statement and every next statement a continuation of the previous one. This happened, then this happened... Ad infinitum. Before today, I could not explain to you why AI articles were so obvious to me, but I think I do now. There is no insight to be gleamed. Pre-LLM, authors generally had intention behind the…

For a limited amount of time I appreciated the level of detail in the article, hoping it would give me more insight, until it exhausted me. I think those two ideas are real takeaways: "Knowing is not doing" and "What can we trust AI to do?". Still, could have been said with a more concise text and maybe a follow up about the details.

The key takeaway pretty much applies to the authoring of the article itself. The LLM knew what all happened but couldn't put it into a readable article.

Re: AI Built a Nuke and Still Lost

#78

Post by teekert »

throwawayqqq11 wrote: post #46
teekert wrote: post #17

Well, the weird thing with nukes is that deterrence only works if you are 100% ready to use them. When the time comes though it would certainly be nice if it turned out to be below 100%. What is winning? Are we a collective or are we individuals? Likely the AI did not get the assignment That "Whatever happens, humans as a race must survive."

Im sure there are some billionaires to find, that finally care about the survival of the white race. /s

Probably

    [f"I'm sure there are some {race} billionaires to find, that finally care about the survival of the {race}." for race in all_races]

Re: AI Built a Nuke and Still Lost

#79

Post by NooneAtAll3 »

pjc50 wrote: post #20

> I now work with governments around the world at the Tony Blair Institute, which means I spend a lot of time in rooms where people ask the same question: what can we actually trust these systems to do? Oh no - we're going to end up with the Starmerbot 3000. Now I've got the joke out of the way, there's at least four interesting lines of inquiry one could take with this blog post: - teaching the AI how to play Civili…

> one guy (on either side!) decides to make hundreds of millions of people worse off out of sheer greed.

I think you greatly low-ball how complex situation is there

Re: AI Built a Nuke and Still Lost

#80

Post by sph »

goatherders wrote: post #59
sph wrote: post #34

Earlier quoted context omitted.

> There is no insight to be gleamed. AI-generated articles are the intellectual equivalent of empty calories. I have just spent the last 10 minutes trying to figure out why someone decided to buy imgui.org, name-squatting an actual project, just to put a slop website on it mildly referencing the original project. It's not even trying to scam you. I keep wondering whether these people that keep polluting the internet…

The answer is in your question. "Empty calories" is a multi trillion dollar business in the food world. It will be the same in the digital world.

McDonalds serves a need: human tend to get hungry.

What need does filling the Internet with AI-generated slop serve? No one is ever going to read those. As I've shown, sometimes they're not even selling anything.

The multi-trillion dollar business is for hyperscalers, but I don't get what slop creators get out of this. What are they spending money for?

Post reply on HN