The future of AI according to thousands of forecasters • Hacker News Board

Re: The future of AI according to thousands of forecasters

#81

Post by TeMPOraL » Wed, Jun 28, 2023, 5:29 PM UTC

nologic01 wrote: post #77

TeMPOraL wrote: post #63

Earlier quoted context omitted.

Counterpoint 1: > Talking about AI (and AGI) as if its some xenomorph lurking somewhere in silicon, waiting for its inevitable escape from its human prison. > AI will not bootstrap itself with some emergent property if somebody spends gazillions of dollars and Watts to estimate petazillions of parameters. Not right now, though recent breakthroughs look worryingly close. But the problem isn't AGI creating itself ex ni…

"AI" overlaps with the more profound and (potentially) far more dangerous problem of the "algorithmization" of society [1]. "Out-of-brain" information technology is something that started long time ago with the agricultural revolution and the first cuneiform tablets. It has been shaping our societies ever since. Once the automation inherent in these constructs is baked-in it doesn't require over-complex black box mod…

Profound, yes. Far more dangerous? Well, it's really the same thing. Automation slotting itself in between people and institutions can be harmful at various scales, but if that automation gets sophisticated enough - that is, once the automation is (equivalent to) powerful AI - we'll be facing an extinction-level threat.

Re: The future of AI according to thousands of forecasters

#82

Post by namaria » Wed, Jun 28, 2023, 10:11 PM UTC

cwkoss wrote: post #48

sublinear wrote: post #43

Earlier quoted context omitted.

> you will be able to do most human tasks with it. You don't need to invent a lot of other stuff to be general purpose. I think this is where most people strongly disagree with you. A probabilistic language model is not good enough to do anything requiring context particularly well.

Agentic uses of GPT can solve the context problem by breaking problems into steps and building prompts to solve those steps. Can't do everything, but GPT+APIs can do a lot.

What in Chomsk's name is "agentic"?

Re: The future of AI according to thousands of forecasters

#83

Post by namaria » Wed, Jun 28, 2023, 10:19 PM UTC

ChatGTP wrote: post #75

kenjackson wrote: post #18

Earlier quoted context omitted.

Depending on who you ask, we will never have it or we have it already. We never had a perfect definition of AGI, but I feel like now the definition has become so murky to be useless.

I kind of wonder if we'll just evolve past the fascination of it, like maybe the problems we need to solve will become so much easier through ML algorithms, we'll just get past it all without the "G" in AGI and become distracted with new things and leave behind being fascinating with computers and chat bots in general. and I know this sounds wild to some, even unthinkable, but I can't imagine people sitting around in…

Some day people will realize that intelligent automaton is an oxymoron. Something designed and built will always operate on rails devised by their creator. Intelligence means the ability of being the creator.

Interestingly, genesis myths usually say god(s) made humans in their image. It's a metaphor for reproduction. We create intelligent beings all the time.

Re: The future of AI according to thousands of forecasters

#84

Post by yunohn » Thu, Jun 29, 2023, 5:40 PM UTC

avionical wrote: post #58

yunohn wrote: post #42

Earlier quoted context omitted.

Yes, but if voters are voting the majority (and probable) decision, that just continues the trend of predicting obvious outcomes. It’s not something to do with understanding domain knowledge and voting based on scientific research.

I wouldn't say AGI in 10 years is predicting an obvious outcome. And the same goes for many other questions on the site.

It will trend towards the major opinion as it gets closer to 10yrs. They haven’t been around long enough.

Re: The future of AI according to thousands of forecasters

#85

Post by _ea1k » Fri, Jun 30, 2023, 12:32 AM UTC

bg24 wrote: post #59

_ea1k wrote: post #34

Earlier quoted context omitted.

Is it surprising that a system that was exclusively trained on emitted language will only do well on language? I don't see how I can extrapolate to anything beyond that. I certainly can't extrapolate that it would be unable to learn other tasks.

Everything is defined in language. As a system is able to learn from the corpus of texts, it also builds reasoning to a large extent. For example, GPT4 can guide you through solving a problem like building a flying and crawling drone for inspecting an attic. Or, watering plants. So this goes beyond language, unless I missed your point.

Karpathy explains it far better than I did in his "State of GPT" talk about a month ago: https://www.youtube.com/watch?v=bZQun8Y4L2A

Yes, language can do that, but books and other texts, on average, leave out a lot of steps that humans have learned to do naturally. They can be prompted to do them, and their answers improve, indicates that there is definitely some level of "reasoning" there.

But it has weird limitations in its current forms. I expect that to improve pretty quickly, tbh.