AI 2027 • Hacker News Board

Re: AI 2027

#141

Post by samr71 » Thu, Apr 03, 2025, 9:39 PM UTC

reducesuffering wrote: post #87

motoxpro wrote: post #29

Earlier quoted context omitted.

That's incredible how much it broadly aligns with what has happened. Especially because it was before ChatGPT.

Will people finally wake up that the AGI X-Risk people have been right and we’re rapidly approaching a really fucking big deal? This forum has been so behind for too long. Sama has been saying this a decade now: “Development of Superhuman machine intelligence is probably the greatest threat to the continued existence of humanity” 2015 https://blog.samaltman.com/machine-intelligence-part-1 Hinton, Ilya, Dario Amodei,…

It's not something you need to worry about.

If we get the Singularity, it's overwhelmingly likely Jesus will return concurrently.

Re: AI 2027

#142

Post by bpodgursky » Thu, Apr 03, 2025, 9:41 PM UTC

superconduct123 wrote: post #115

Why are the biggest AI predictions always made by people who aren't deep in the tech side of it? Or actually trying to use the models day-to-day...

Because you can't be a full time blogger and also a full time engineer. Both take all your time, even ignoring time taken to build talent. There is simply a tradeoff of what you do with your life.

There are engineers with AI predictions, but you aren't reading them, because building an audience like Scott Alexander takes decades.

Re: AI 2027

#143

Post by torginus » Thu, Apr 03, 2025, 9:44 PM UTC

Much has been made in its article about autonomous agents ability to do research via browsing the web - the web is 90% garbage by weight (including articles on certain specialist topics).

And it shows. When I used GPT's deep research to research the topic, it generated a shallow and largely incorrect summary of the issue, owning mostly to its inability to find quality material, instead it ended up going for places like Wikipedia, and random infomercial listicles found on Google.

I have a trusty Electronics textbook written in the 80s, I'm sure generating a similarly accurate, correct and deep analysis on circuit design using only Google to help would be 1000x harder than sitting down and working through that book and understanding it.

Re: AI 2027

#144

Post by programd » Thu, Apr 03, 2025, 9:44 PM UTC

kubb wrote: post #44

mitthrowaway2 wrote: post #14

Earlier quoted context omitted.

What's an example of an intellectual task that you don't think AI will be capable of by 2027?

It won't be able to write a compelling novel, or build a software system solving a real-world problem, or operate heavy machinery, create a sprite sheet or 3d models, design a building or teach. Long term planning and execution and operating in the physical world is not within reach. Slight variations of known problems should be possible (as long as the size of the solution is small enough).

Does a fighter jet count as "heavy machinery"?

https://apnews.com/article/artificial-intelligence-fighter-j...

Re: AI 2027

#145

Post by elicksaur » Thu, Apr 03, 2025, 9:45 PM UTC

crazystar wrote: post #102

Jun8 wrote: post #96

ACT post where Scott Alexander provides some additional info: https://www.astralcodexten.com/p/introducing-ai-2027 . Manifold currently predicts 30%: https://manifold.markets/IsaacKing/ai-2027-reports-predictio...

47% now soo a coin toss

Note the market resolves by:

> Resolution will be via a poll of Manifold moderators. If they're split on the issue, with anywhere from 30% to 70% YES votes, it'll resolve to the proportion of YES votes.

So you should really read it as “Will >30% of Manifold moderators in 2027 think the ‘predictions seem to have been roughly correct up until that point’?”

Re: AI 2027

#146

Post by goatlover » Thu, Apr 03, 2025, 9:46 PM UTC

reducesuffering wrote: post #87

motoxpro wrote: post #29

Earlier quoted context omitted.

That's incredible how much it broadly aligns with what has happened. Especially because it was before ChatGPT.

Will people finally wake up that the AGI X-Risk people have been right and we’re rapidly approaching a really fucking big deal? This forum has been so behind for too long. Sama has been saying this a decade now: “Development of Superhuman machine intelligence is probably the greatest threat to the continued existence of humanity” 2015 https://blog.samaltman.com/machine-intelligence-part-1 Hinton, Ilya, Dario Amodei,…

> "Development of Superhuman machine intelligence is probably the greatest threat to the continued existence of humanity”

If that's really true, why is there such a big push to rapidly improve AI? I'm guessing OpenAI, Google, Anthropic, Apple, Meta, Boston Dynamics don't really believe this. They believe AI will make them billions. What is OpenAI's definition of AGI? A model that makes $100 billion?

Re: AI 2027

#147

Post by pixl97 » Thu, Apr 03, 2025, 9:47 PM UTC

barbarr wrote: post #82

moab wrote: post #13

> "OpenBrain (the leading US AI project) builds AI agents that are good enough to dramatically accelerate their research. The humans, who up until very recently had been the best AI researchers on the planet, sit back and watch the AIs do their jobs, making better and better AI systems." I'm not sure what gives the authors the confidence to predict such statements. Wishful thinking? Worst-case paranoia? I agree that…

It also ignores the possibility of plateau... maybe there's a maximum amount of intelligence that matter can support, and it doesn't scale up with copies or speed.

Eh, these mathematics still don't work out in humans favor...

Lets say intelligence caps out at the maximum smartest person that's ever lived. Well, the first thing we'd attempt to do is build machines up to that limit that 99.99999 percent of us would never get close to. Moreso the thinking parts of humans is only around 2 pounds of mush in side of our heads. On top of that you don't have to grow them for 18 years first before they start outputting something useful. That and they won't need sleep. Oh and you can feed them with solar panels. And they won't be getting distracted by that super sleek server rack across the aisle.

We do know 'hive' or societal intelligence does scale over time especially with integration with tooling. The amount of knowledge we have and the means of which we can apply it simply dwarf previous generations.

Re: AI 2027

#148

Post by jaccola » Thu, Apr 03, 2025, 9:47 PM UTC

hiq wrote: post #74

amarcheschi wrote: post #12

I just spent some time trying to make claude and gemini make a violin plot of some polar dataframe. I've never used it and it's just for prototyping so i just went "apply a log to the values and make a violin plot of this polars dataframe". ANd had to iterate with them for 4/5 times each. Gemini got it right but then used deprecated methods I might be doing llm wrong, but i just can't get how people might actually do…

> had to iterate with them for 4/5 times each. Gemini got it right but then used deprecated methods How hard would it be to automate these iterations? How hard would it be to automatically check and improve the code to avoid deprecated methods? I agree that most products are still underwhelming, but that doesn't mean that the underlying tech is not already enough to deliver better LLM-based products. Lately I've been…

How hard can it be to create a universal "correctness" checker? Pretty damn hard!

Our notion of "correct" for most things is basically derived from a very long training run on reality with the loss function being for how long a gene propagated.

Re: AI 2027

#149

Post by achierius » Thu, Apr 03, 2025, 9:47 PM UTC

hn_throwaway_99 wrote: post #129

reducesuffering wrote: post #87

Earlier quoted context omitted.

Will people finally wake up that the AGI X-Risk people have been right and we’re rapidly approaching a really fucking big deal? This forum has been so behind for too long. Sama has been saying this a decade now: “Development of Superhuman machine intelligence is probably the greatest threat to the continued existence of humanity” 2015 https://blog.samaltman.com/machine-intelligence-part-1 Hinton, Ilya, Dario Amodei,…

> Will people finally wake up that the AGI X-Risk people have been right and we’re rapidly approaching a really fucking big deal? OK, say I totally believe this. What, pray tell, are we supposed to do about it? Don't you at least see the irony of quoting Sama's dire warnings about the development of AI, without at least mentioning that he is at the absolute forefront of the push to build this technology that can dest…

Political organization to force a stop to ongoing research? Protest outside OAI HQ? There are lots of thing we could, and many of us would, do if more people were actually convinced their life were in danger.

Re: AI 2027

#150

Post by elicksaur » Thu, Apr 03, 2025, 9:50 PM UTC

madethisnow wrote: post #95

samth wrote: post #93

Earlier quoted context omitted.

I think it's not holding up that well outside of predictions about AI research itself. In particular, he makes a lot of predictions about AI impact on persuasion, propaganda, the information environment, etc that have not happened.

something you can't know

This doesn’t seem like a great way to reason about the predictions.

For something like this, saying “There is no evidence showing it” is a good enough refutation.

Counterpointing that “Well, there could be a lot of this going on, but it is in secret.” - that could be a justification for any kooky theory out there. Bigfoot, UFOs, ghosts. Maybe AI has already replaced all of us and we’re Cylons. Something we couldn’t know.

The predictions are specific enough that they are falsifiable, so they should stand or fall based on the clear material evidence supporting or contradicting them.