DeepMind says reinforcement learning is ‘enough’ to reach general AI

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by TeMPOraL on Thursday June 10, 2021 @04:17AM (#27458270)

I'll happily bet fake internet points instead:

https://www.metaculus.com/questions/create/

EDIT: bunch of other related predictions currently open:

https://www.metaculus.com/questions/?order_by=-activity&sear...

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by marcescence on Thursday June 10, 2021 @04:36AM (#27458392)

A quote from datscilly, the top forecaster on metaculus:

>AGI may never happen, but the chance of that is small enough that adjusting for that here will not make a big difference (I put ~10% that AGI will not happen for 500 years or more, but it already matches that distribution quite well).[1]

[1]:https://www.lesswrong.com/posts/hQysqfSEzciRazx8k/forecastin...

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by okprod on Thursday June 10, 2021 @10:21AM (#27461104)

while Deep Mind has a huge financial temptation to sell snake oil

I don't know, isn't the DeepMind founder that guy in the Go documentary? I read about him after watching the doc and he seemed to be pretty cautious about taking in investment, and he didn't seem the type to try to cash out.

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by ocdtrekkie on Thursday June 10, 2021 @10:52AM (#27461430)

He already cashed out, he sold to Google. And over the years Google has ramped up the pressure for DeepMind to deliver financial returns. (I recall when Google tried to stick DeepMind's branding on GCP, Watson-style, so it would sell better, and at the time, DeepMind was able to decline.)

Eventually Google will give them the option to deliver financial success or be shut down.

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by for_i_in_range on Thursday June 10, 2021 @03:42AM (#27458013)

“Show me the incentives I’ll show you the outcome.”

Google make money, Google Bad.

Deep Mind owned by Google, Deep Mind bad!

The above conclusion stands trite.

Perhaps the inverse is true.

Google and Deep Mind, if correct, could be hurting themselves more than helping themselves.

Why? Creating a future species who’s too smart to click on ads, and too smart to remain subject to its whims, doesn’t sound like it’d be good for quarterly profits...

There’s also the emotional incentive for humans to confirm their own beliefs about humanity being special.

If Google/Deep Mind knows this, yet publishes research anyway in the spirit of truth, why, what they’re doing may be considered heroic.

Two sides of the coin here.

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by rocgf on Thursday June 10, 2021 @03:58AM (#27458136)

No offence, but I think you are extremely wrong.

Creating an AGI is the endgame for everything. Who cares about ads when you have an AI that can learn to do anything and improve upon itself continuously?

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by zcw100 on Thursday June 10, 2021 @09:13AM (#27460308)

Really? I created two GI's and it wasn't very hard and was actually quite fun. Training them is a bit of a pain though. I'm willing to bet that based on total calories consumed they are amazingly efficient compared to their hypothetical AGI counterparts.

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by FredPret on Thursday June 10, 2021 @10:53AM (#27461442)

Yes but can they:

- live forever

- grow their own mental capabilities exponentially over that unlimited lifespan

- turn themselves into universe-eating von Neumann probes

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by rsj_hn on Thursday June 10, 2021 @01:36PM (#27463298)

Nothing can

- live forever

- grow exponentially forever

- "eat the universe" (I know, the last point was sci-fi gibberish)

In fact, humans are already pretty good at reproducing themselves and have managed to travel to space, and have exhibited finite periods of exponential knowledge growth combined with periods of collapse, as nothing grows exponentially forever.

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by mensetmanusman on Thursday June 10, 2021 @02:30PM (#27463857)

My three-year-old said yes to all questions

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by croon on Thursday June 10, 2021 @04:04AM (#27458188)

You don't know if an AGI will agree with your profit motives.

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by rocgf on Thursday June 10, 2021 @06:56AM (#27459226)

There is a huge assumption baked into your comment and I do not agree with it.

AGI does not necessarily require for it to be conscious or throw tantrums about its creators' purpose. AGI just means that it's an intelligence that can be thrown at any problem, not just a particular game or task, similar to how humans can specialize in CS or playing the violin.

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by croon on Thursday June 10, 2021 @07:12AM (#27459332)

Sure, it was somewhat tongue in cheek, but not entirely.

There is a semi-established definition that does include what I referred to:

> AGI can also be referred to as strong AI,[2][3][4] full AI,[5] or general intelligent action.[6] Some academic sources reserve the term "strong AI" for computer programs that can experience sentience, self-awareness and consciousness.[7]

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by kordlessagain on Thursday June 10, 2021 @11:43AM (#27462027)

That’s not going to happen until they have bodies.

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by for_i_in_range on Thursday June 10, 2021 @04:03AM (#27458177)

“Who cares?” Well, the people who need to pay the people developing the endgame for everything you speak of.

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by rocgf on Thursday June 10, 2021 @06:59AM (#27459252)

I'm sorry, but this makes no sense to me.

The people paying for the development of the AGI can mean many things - the Google customers/users, Alphabet as a company, the executives throwing money at the problem?

Either way, I don't really get your point. Your initial post was about how it is counterintuitive for Google to allocate funds for an AGI, since it makes money out of ads. These are not mutually exclusive, you can have both, but my point is that if you develop an AGI, then you can pretty much "conquer" the world and revenue from ads becomes irrelevant.

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by for_i_in_range on Thursday June 10, 2021 @03:17PM (#27464451)

How do you think they can conquer the world? How do you foresee governments not restricting a private company’s new powerful tool?

[ Parent | on HN ]