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

Horses: AI progress is steady. Human equivalence is sudden

andyljones.com

Post reply on HN
Print view View original

41–50 of 663 posts

Re: Horses: AI progress is steady. Human equivalence is sudden

#41

Post by shermantanktop »

adventured wrote: post #16

It's astounding how subtly anti-AI HN has become over the past year, as the models keep getting better and better. It's now pervasive across nearly every AI thread here. As the potential of AI technical agents has gone from an interesting discussion to extraordinarily obvious as to what the outcome is going to be, HN has comically shifted negative in tone on AI. They doth protest too much. I think it's a very clear c…

It's not subtle.

But the temptation of easy ideas cuts both ways. "Oldsters hate change" is a blanket dismissal, and there are legitimate concerns in that body of comments.

Re: Horses: AI progress is steady. Human equivalence is sudden

#42

Post by trollbridge »

echelon wrote: post #29

> And not very long after, 93 per cent of those horses had disappeared. > I very much hope we'll get the two decades that horses did. > But looking at how fast Claude is automating my job, I think we're getting a lot less. This "our company is onto the discovery that will put you all out of work (or kill you?)" rhetoric makes me angry. Something this powerful and disruptive (if it is such) doesn't need to be owned or…

At this point "we're going to make all office work obsolete" feels more like a marketing technique than anything actually connected to reality. It's sort of like how Coca-Cola implies that drinking their stuff will make you popular and well-liked by other attractive, popular people.

Meanwhile, my own office is buried in busywork that there are currently no AI tools on the market that will do the work for us, and AI entering a space sometimes increases busywork workloads. For example, when writing descriptions of publications or listings for online sales, we have to put more effort now into not sounding like it was AI-generated or we will lose sales. The AI tools for writing descriptions / generating listings are not very helpful either. (An inaccurate listing/description is a nightmare.)

I was able to help set up a client with AI tools to help him generate basically a faux website in a few hours that has lots of nice graphic design, images, etc. so that his new venture looks like a real company. Well, except for the "About Us" page that hallucinated an executive team plus a staff of half a dozen employees. So I guess work like that does get done faster now.

Re: Horses: AI progress is steady. Human equivalence is sudden

#43

Post by danpalmer »

personjerry wrote: post #3

I think it's a cool perspective, but the not-so-hidden assumption is that for any given domain, the efficiency asymptote peaks well above the alternative. And that really is the entire question at this point: Which domains will AI win in by a sufficient margin to be worth it?

> the not-so-hidden assumption is that for any given domain, the efficiency asymptote peaks well above the alternative

This is an assumption for the best-case scenario, but I think you could also just take the marginal case. Steady progress builds until you get past the state of the art system, and then the switch becomes easy to justify.

Re: Horses: AI progress is steady. Human equivalence is sudden

#44

Post by burroisolator »

falcor84 wrote: post #19
burroisolator wrote: post #6

"In 1920, there were 25 million horses in the United States, 25 million horses totally ambivalent to two hundred years of progress in mechanical engines. And not very long after, 93 per cent of those horses had disappeared. I very much hope we'll get the two decades that horses did." I'm reminded of the idiom "be careful what you wish for, as you might just get it." Rapid technogical change has historically lead to p…

My reading of tfa is exactly that - the author is hoping that we'll have at least a generation or so to adapt, like horses did, but is concerned that it might be significantly more rapid.

"You're absolutely right!" Thanks for pointing it out. I was expecting that kind of perspective when the author brought up horses, but found the conclusion to be odd. Turns out it was just my reading of it.

Re: Horses: AI progress is steady. Human equivalence is sudden

#45

Post by Lerc »

adventured wrote: post #16

It's astounding how subtly anti-AI HN has become over the past year, as the models keep getting better and better. It's now pervasive across nearly every AI thread here. As the potential of AI technical agents has gone from an interesting discussion to extraordinarily obvious as to what the outcome is going to be, HN has comically shifted negative in tone on AI. They doth protest too much. I think it's a very clear c…

>It's astounding how subtly anti-AI HN has become over the past year, as the models keep getting better and better. It's now pervasive across nearly every AI thread here.

I don't think you can characterise it as a sentiment of the community as a whole. While every AI thread seems to have it's share of AI detractors, the usernames of the posters are becoming familiar. I think it might be more accurate to say that there is a very active subset of users with that opinion.

This might hold true for the discourse in the wider community. You see a lot of coverage about artists outraged by AI, but when I speak to artists they have a much more moderate opinion. Cautious, but intrigued. A good number of them are looking forward to a world that embraces more ambitious creativity. If AI can replicate things within a standard deviation of the mean, the abundance of that content there will create an appetite for something further out.

Re: Horses: AI progress is steady. Human equivalence is sudden

#47

Post by whycombinetor »

jacobsenscott wrote: post #28
raincole wrote: post #14

Earlier quoted context omitted.

What example do you need? In every single benchmark AI is getting better and better. Before someone says "but benchmark doesn't reflect real world..." please name what metric you think is meaningful if not benchmark. Token consumption? OpenAI/Anthropic revenue?

Whenever I try and use a "state of the art" LLM to generate code it takes longer to get a worse result than if I just wrote the code myself from the start. That's the experience of every good dev I know. So that's my benchmark. AI benchmarks are BS marketing gimmicks designed to give the appearance of progress - there are tremendous perverse financial incentives. This will never change because you can only use an LLM…

Third party benchmarks like terminalbench exist.

W.r.t code changes especially small ones (say 50 lines spread across 5 files), if you can't get an agent to make nearly exactly the code changes you want, just faster than you, that's a you problem at this point. If it maybe would take you 15 minutes, grok-code-fast-1 can do it in 2.

Re: Horses: AI progress is steady. Human equivalence is sudden

#48

Post by trollbridge »

jacobsenscott wrote: post #28
raincole wrote: post #14

Earlier quoted context omitted.

What example do you need? In every single benchmark AI is getting better and better. Before someone says "but benchmark doesn't reflect real world..." please name what metric you think is meaningful if not benchmark. Token consumption? OpenAI/Anthropic revenue?

Whenever I try and use a "state of the art" LLM to generate code it takes longer to get a worse result than if I just wrote the code myself from the start. That's the experience of every good dev I know. So that's my benchmark. AI benchmarks are BS marketing gimmicks designed to give the appearance of progress - there are tremendous perverse financial incentives. This will never change because you can only use an LLM…

Right. With careful use of AIs, I can use it to gather information to help me make better designs (like giving me summaries of the current best available frameworks or libraries to choose for a given project), but as far as just generating an architecture and then generating the code and devops and so on for that? It's just not there, unless you're creating an app that effectively already exists, like some basic CRUD app.

If you're creating basic CRUDs, what on earth are you doing? That kind of thing should have been automated a long time ago.

Re: Horses: AI progress is steady. Human equivalence is sudden

#49

Post by stego-tech »

1970-01-01 wrote: post #24

How about we stop trying the analogy clothing on and just tell it like it is? AI is unlike any other technology to date. Just like predicting the weather, we don't know what it will be like in 20 months. Everything is a guesstimate.

This is the correct take. We all have that "Come to Jesus" moment eventually, where something blows our minds so profoundly that we believe anything is possible in the immediate future. I respect that, it's a great take to have and promotes a lot of discussion, but now more than ever we need concretes and definitives instead of hype machines and their adjacent counterparts.

Too much is on the line here regardless of what ultimately ends up being true or just hype.

Re: Horses: AI progress is steady. Human equivalence is sudden

#50

Post by billisonline »

An engine performs a simple mechanical operation. Chess is a closed domain. An AI that could fully automate the job of these new hires, rather than doing RAG over a knowledge base to help onboard them, would have to be far more general than either an engine or a chessbot. This generality used to be foregrounded by the term "AGI." But six months to a year ago when the rate of change in LLMs slowed down, and those exciting exponentials started to look more like plateauing S-curves, executives conveniently stopped using the term "AGI," preferring weasel-words like "transformative AI" instead.

I'm still waiting for something that can learn and adapt itself to new tasks as well as humans can, and something that can reason symbolically about novel domains as well as we can. I've seen about enough from LLMs, and I agree with the critique that som type of breakthrough neuro-symbolic reasoning architecture will be needed. The article is right about one thing: in that moment AI will overtake us suddenly! But I doubt we will make linear progress toward that goal. It could happen in one year, five, ten, fifty, or never. In 2023 I was deeply concerned about being made obsolete by AI, but now I sleep pretty soundly knowing the status quo will more or less continue until Judgment Day, which I can't influence anyway.

Post reply on HN