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

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by londons_explore on Thursday June 10, 2021 @03:13AM (#27457810)

Nature has a massive incentive to make good use of energy from light through photosynthesis. Billions of plants compete, and whoever can get most out of the sun will win out.

Yet manmade solar cells are more efficient by nearly all measures.

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by Clewza313 on Thursday June 10, 2021 @03:17AM (#27457840)

Except that manmade solar cells are pretty bad at repairing or replicating themselves.

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by rich_sasha on Thursday June 10, 2021 @04:40AM (#27458413)

Or growing out of literally nothing but dirt and water.

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by 21eleven on Thursday June 10, 2021 @10:39AM (#27461298)

And air. That's what is crazy about plants, their carbon comes from the CO2 in the air.

Also if someone loses weight, most of the carbon that made up their fat leaves the body as breath.

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by khimaros on Thursday June 10, 2021 @08:46PM (#27467982)

and are inedible

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by khimaros on Thursday June 10, 2021 @08:47PM (#27467989)

and are a damned eye-sore

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by nindalf on Thursday June 10, 2021 @10:44AM (#27461344)

The solar cells get some organic life form to assist in their reproduction phase. That's pretty efficient too.

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by EGreg on Thursday June 10, 2021 @10:59AM (#27461514)

Why are these just-so stories believed so much?

Just because plants compete on some limited level doesn’t mean that a particular plant organism “winning” means becoming the most efficient converter of sunlight.

Is everyone’s memory like those people who can remember every detail? Why not? If you’re immediately planning to make up a just-so explanation on the spot that has the requisite but unproven claim about increasing the genetic fitness function, that is the problem with evolutionary explanations. It’s not science if you just make stuff up and give it the same amount of credibility as something that has been tested and proven. You can take any trait and spin stories about why it is the way it is, and then expect somehow that some metric has to be maximized because of your unproven theory.

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by bildung on Thursday June 10, 2021 @03:18AM (#27457850)

> Yet manmade solar cells are more efficient by nearly all measures.

Only because we cheated, though: Houses can't sponantously grow more cells in place when more energy is needed.

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by maaaaattttt on Thursday June 10, 2021 @03:44AM (#27458028)

On a half-jokingly note, they can, their humans buy them and put them where needed. An alien observer in space would see some houses spontaneously growing solar cells on their roofs.

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by FredPret on Thursday June 10, 2021 @10:40AM (#27461307)

This is even more interesting if you think of all human artifacts as being equivalent to anthills and beaver dams.

1) Trees are natural and trees create leaves with a solar efficiency of x

2) Humans are natural and we create solar panels with efficiency x + y

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by chriswarbo on Thursday June 10, 2021 @11:34AM (#27461923)

This is Dawkins' idea of "extended phenotype". Normally a gene's phenotype refers to its effects on the body of an individual organism posessing that gene, like hair colour or immune response.

A gene's extended phenotype includes effects external to particular organisms, like nests, deforestation, changes to the chemical makeup of the atmosphere, etc.

[ Parent | on HN ]

Re:DeepMind says reinforcement learning is ‘enough’ to reach... by idiotsecant on Thursday June 10, 2021 @11:05AM (#27461570)

nature only has an incentive to increase efficiency when that increase in efficiency results in increased chance of producing gene copies.

Nature is full of examples that are 'good enough' while balancing other competing constraints. Evolution doesn't create organisms optimized for efficiency - it creates organisms optimized for reproduction. The two are not always the same.

[ Parent | on HN ]