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

LLMs are the ultimate demoware

blog.charliemeyer.co

Post reply on HN
Print view View original

111–120 of 128 posts

Re: LLMs are the ultimate demoware

#111

Post by rsynnott »

js8 wrote: post #9
dpflan wrote: post #2

Yes, AI allows for exquisite demos, demos that tantalize the audience into thinking of the infinite potential of the technology, that stunning vision expands and expands until the universe of potential overwhelms the dreamer into a state of terminal fantasy. So it is always a solution looking for a problem. There are cases where the two meet more realistically and a valuable impactful company develops it. The fact it…

LLM models have evolved to autonomously convince humans that they're useful. They're the ultimate memetic parasites.

... I mean, it's not evolution. These things have people guiding them. Note the whole 'agreeability' controversy. That one's is a bit like cigarette companies back in the day optimising their products for addictiveness; do you do the right thing, or the thing that makes people buy your product more?

Re: LLMs are the ultimate demoware

#113

Post by rsynnott »

lm28469 wrote: post #70
slaterbug wrote: post #5

Earlier quoted context omitted.

What evidence is there that AGI will come “soon”?

They can "feel it", like people "felt" we'd have commercial space flight "soon" after we put people on the moon, it's all delusion and wishful thinking.

It's worse than that, really, because there was at least a fairly obvious _path_ there, even if the economics were, to say the least, shaky. For AGI... not so much.

Re: LLMs are the ultimate demoware

#114

Post by rsynnott »

eitland wrote: post #4

Again and again people keep saying this while many of us keep using LLMs to create value.

The trouble is that peoples' self-evaluation of things that they believe are helping them is generally poor, and there's, at best, weak and conflicting evidence which is _not_ based on polling users.

In particular, "producing stuff" is not necessarily "creating value"; some stuff has _negative_ value.

Re: LLMs are the ultimate demoware

#115

Post by simonw »

rsynnott wrote: post #110
empath75 wrote: post #10

Earlier quoted context omitted.

It seems very hard to maintain the belief that LLMs are useless in the face of the fact that millions of people are using them. It's very much "nobody goes there anymore, it's too crowded"

On that basis you'll also be adopting TCM, and homeopathy, and dowsing, and all major religions simultaneously, and all major fad diets simultaneously? Like, "lots of people like this thing and think it is helping them" is not terribly strong evidence that it is actually helping them. It's not a good argument.

How about "lots of people like this thing" where many of those people are credible professionals who I have respected from LONG before they started using LLMs?

Re: LLMs are the ultimate demoware

#116

Post by rsynnott »

simonw wrote: post #115
rsynnott wrote: post #110

Earlier quoted context omitted.

On that basis you'll also be adopting TCM, and homeopathy, and dowsing, and all major religions simultaneously, and all major fad diets simultaneously? Like, "lots of people like this thing and think it is helping them" is not terribly strong evidence that it is actually helping them. It's not a good argument.

How about "lots of people like this thing" where many of those people are credible professionals who I have respected from LONG before they started using LLMs?

Again, it's not _hugely_ strong evidence. Linus Pauling won _two_ Nobel prizes, was unquestionably brilliant... but that doesn't mean I'm going to start megadosing vitamin C anytime soon (https://en.wikipedia.org/wiki/Linus_Pauling#Medical_research...).

(See also Newton and alchemy, and the list goes on.)

Re: LLMs are the ultimate demoware

#117

Post by simonw »

rsynnott wrote: post #116
simonw wrote: post #115

Earlier quoted context omitted.

How about "lots of people like this thing" where many of those people are credible professionals who I have respected from LONG before they started using LLMs?

Again, it's not _hugely_ strong evidence. Linus Pauling won _two_ Nobel prizes, was unquestionably brilliant... but that doesn't mean I'm going to start megadosing vitamin C anytime soon ( https://en.wikipedia.org/wiki/Linus_Pauling#Medical_research... ). (See also Newton and alchemy, and the list goes on.)

The experts I respect on this are people who's expertise is in software development, so when they say "this stuff helps me do what I do better and faster" I trust them more than if they said "I've started megadosing vitamin C and it's amazing".

It helps that what they're discovering matches my own personal experience as well.

Re: LLMs are the ultimate demoware

#118

Post by simonw »

simonw wrote: post #68
pjc50 wrote: post #35

Earlier quoted context omitted.

What's the acceptance rate on such PRs?

Add is:merged to see. For Copilot I got 151,000 out of 247,000 = 61% For Claude 124,000 / 147,000 = 84% For Codex 1.7m / 2m = 85%

... I just found out there's an existing repo and site that's been running these kinds of searches for a while: https://prarena.ai/ and https://github.com/aavetis/PRarena

Re: LLMs are the ultimate demoware

#119

Post by kasey_junk »

tptacek wrote: post #89
benterix wrote: post #88

Earlier quoted context omitted.

While I agree, on an unrelated note - I knew I know your nick from somewhere... And then I realized[0]. [0] https://ludic.mataroa.blog/blog/contra-ptaceks-terrible-arti...

I had a conversation with that person a couple weeks ago. They're nice. I think we both would tweak (if just a little bit) how we presented our articles with the benefit of hindsight. For the record, I'm a systems programmer and a security person and I don't work for an AI company (you can Six Degrees of Sam Altman any startup to AI now if you want to make the claim, but if you try I'm just going to say "Sir, This Is…

I on the other hand am the mental case that drinks rocket fuel.

Re: LLMs are the ultimate demoware

#120

Post by dsr_ »

slaterbug wrote: post #25
dsr_ wrote: post #11

Earlier quoted context omitted.

Or "ever"? (I'm not denying the possiblity. I'm proclaiming a lack of evidence.)

I’ve been daydreaming lately about what the fundamental limits of “intelligence” could be, something like the concept of computability but for AI, or even biological brains. Though I will say, surely the existence of the human brain (which by definition is general intelligence), suggests that creating AGI is fundamentally possible?

Sure, it's possible - as you say, we have an existence proof. We don't know how to do it any other way, though. None of the people who claim that they or somebody else is on the trail has produced any evidence that they are correct.
Post reply on HN