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

On being wrong about AI

scottaaronson.blog

Post reply on HN
Print view View original

61–70 of 79 posts

Re: On being wrong about AI

#61

Post by fragmede »

sigmoid10 wrote: post #19

This article is absolutely pointless. The statement was from 2009. Back then, neither side had any real evidence in favour of their argument. Deep learning and AI were still purely academic experiments, with zero practicality. And it had been like this for half a century. Back then, even the best models couldn't tell cats from dogs. Today every high schooler with a laptop can solve that problem thanks to abundant har…

If people keep bringing up something you wrote in 2009 and called you a dumbass for saying what you said 14 years ago, wouldn't you put out a statement saying they're being dumbasses today?

Re: On being wrong about AI

#62

Post by treprinum »

Jensson wrote: post #8

> He seems to have thought someone was going to discover a revolutionary “key” to AI. That didn’t happen; you might say I was right to be skeptical of it. Transformer architecture happened, that is key for neural networks to build up context dependent reasoning, we wouldn't have gotten to where we are today without it. Without it scaling neural networks doesn't add much value, since without being able to separate out…

Transformers are synonymous with scaling as it was the first model that kept improving both by adding more and more layers and throwing more training data at it. All the previous models like CNNs, ResNet, LSTMs etc. hit some depth/training size where no progress was happening anymore as gradient was unreliably flowing down.

Re: On being wrong about AI

#63

Post by mark_l_watson »

piva00 wrote: post #49
mark_l_watson wrote: post #36

Earlier quoted context omitted.

I think you misunderstood what I was saying, so I will try to be more clear: I was describing myself as an observer and participant in different waves of AI tech. I was not admitting to being wrong, rather, just someone who was lucky enough to be a small part of AI. What I don’t know right now: if AI will help solve difficult societal/energy/medical/scientific problems and propel humanity into a better future (I bet…

In my purely layman view of someone on the outskirts of AI development, meaning that I knew about the techniques from a basic concept view, saw the rise of ML and NN into the mainstream software development the past 15-20 years (and not from an academic background), and now seeing the latest generative AIs and LLMs. My biggest anxiety with AI development rapidly increasing in pace and capabilities is that we don't wo…

I worry about this also: how does society function when only half the workers are required? We will need some sort of minimal social safety net that still allows people to do some work and get paid.

I view future AIs as being partners, working with people. That said, much less human labor will be required.

Not really addressing your good comments, but as an aside, I really hope that open AI models “win” rather than opaque AI systems owned and run by 3 or 4 giant tech companies.

Re: On being wrong about AI

#64

Post by TomaszZielinski »

mark_l_watson wrote: post #10

Scott shouldn’t be hard on himself at all for opinions stated 15 years ago. Even over short time spans of a few years, I change my perspective quite a bit. In the 1980s I was lucky enough to stumble into the opportunity of serving on a DARPA neural networks tools panel for a year and getting lucky applying a simple backprop model to a show-stopper problem for a bomb detector my company designed and built for the FAA.…

> Scott shouldn’t be hard on himself at all for opinions stated 15 years ago. Even over short time spans of a few years, I change my perspective quite a bit.

I've came here to write a similar thing. Fifteen years is a long time, and just as you wrote, I also tend to change my perspective and opinions quite a bit (and sometimes even quite a lot) over much shorter time spans.

And I'm not even talking about tech--but about life in general. In fact, I remember some of my opinions from 5, 10, 15 year ago, and some of them are cringeworthy (to put it rather mildly) to my present self...

Re: On being wrong about AI

#66

Post by Kuinox »

hotpotamus wrote: post #60
Kuinox wrote: post #53

Earlier quoted context omitted.

Supersonic passenger jets have another problem, the Concorde demonstated it: it made so much noise countries forbid it from flying over their land.

Then the real innovation that we’ve lacked since the Concorde is quiet (and probably economical) supersonic flight.

It's a long way until commercial usage since even prototypes didn't flew yet. https://en.wikipedia.org/wiki/Lockheed_Martin_X-59_Quesst

Re: On being wrong about AI

#67

Post by aleph_minus_one »

its-summertime wrote: post #20
blueflow wrote: post #3

> civilization-altering trajectory On which planet? Because i don't see any of it. AI should have made my console monkey job obsolete, by yesterday.

Is it impacting hiring decisions? Changes in government policy? etc. If not already, how about in the next 5 years? A majority of the use-cases are not going to be publicly attributed to AI, that however does not mean it isn't having an effect.

> Is it impacting hiring decisions? [...] If not already, how about in the next 5 years?

At least in the field of activity that I work in: No. I honestly don't have the slightest idea how AI could be really helpful for hiring decisions. The only "useful" usage of AI in hiring is to serve as a scapegoat for covering your a... if the hiring decision turned out to be a bad idea ("but the AI said ...").

Re: On being wrong about AI

#68

Post by sigmoid10 »

T-A wrote: post #58
sigmoid10 wrote: post #19

This article is absolutely pointless. The statement was from 2009. Back then, neither side had any real evidence in favour of their argument. Deep learning and AI were still purely academic experiments, with zero practicality. And it had been like this for half a century. Back then, even the best models couldn't tell cats from dogs. Today every high schooler with a laptop can solve that problem thanks to abundant har…

> This article is absolutely pointless. Not at all, but maybe you didn't read far enough to get to it. > The statement was from 2009. Back then, neither side had any real evidence in favour of their argument. Deep learning and AI were still purely academic experiments, with zero practicality. That's exactly what the article says: I was explaining that there was no trend you could knowably, reliably project into the f…

>maybe you didn't read far enough to get to it.

Perhaps you didn't read my comment far enough to get it. The point is - essentially - you can't trust experts in any field when they start arguing beliefs instead of science. It has nothing to do with confidence in technological pacing or lack thereof. It seems the author has learned nothing of his real mistake and would be prone to repeating it in the future.

Re: On being wrong about AI

#69

Post by sigmoid10 »

fragmede wrote: post #61
sigmoid10 wrote: post #19

This article is absolutely pointless. The statement was from 2009. Back then, neither side had any real evidence in favour of their argument. Deep learning and AI were still purely academic experiments, with zero practicality. And it had been like this for half a century. Back then, even the best models couldn't tell cats from dogs. Today every high schooler with a laptop can solve that problem thanks to abundant har…

If people keep bringing up something you wrote in 2009 and called you a dumbass for saying what you said 14 years ago, wouldn't you put out a statement saying they're being dumbasses today?

That's the thing people are not getting: The author was a dumbass. As was his counterpart. But not for the reasons that were outlined.

Re: On being wrong about AI

#70

Post by gilbetron »

aleph_minus_one wrote: post #7

"For that matter, hardly anyone predicted the staggering, civilization-altering trajectory of neural network performance from roughly 2012 onwards." Whether the trajectory will be civilization-altering is yet to be seen. Let me put it this way: I have serious doubts.

It already is altering civilization, so I don't understand your doubts. I work for a large unicorn tech and in a single year we went from execs saying "we'll look into this AI thing but we assume it is the usual hype" to creating multiple, successful AI products and encouraging developers to use AI software tools because of the recognition of a significant benefit to the business when they are used. We are at the very start of a logistic curve, and that initial exponential rate can seem like nothing. In a few years, we'll really start to feel it and in a decade or so it the change will become a blur, more like an instantaneous shift to another civilization.
Post reply on HN