Researchers Teach Human Brain Cells In a Dish To Play 'Pong' (futurism.com) 44
Slashdot reader Hmmmmmm quotes a report from Futurism:
Researchers at the biotechnology startup Cortical Labs have created "mini-brains" consisting of 800,000 to one million living human brain cells in a petri dish, New Scientist reports.
The cells are placed on top of a microelectrode array that analyzes the neural activity... To teach the mini-brains the game, the team created a simplified version of "Pong" with no opponent. A signal is sent to either the right or left of the array to indicate where the ball is, and the neurons from the brain cells send signals back to move the paddle...
Brett Kagan, chief scientific officer at Cortical Labs and research lead of the project, said that while the mini-brains can't play the game as well as a human, they do learn faster than some AIs.
"The amazing aspect is how quickly it learns, in five minutes, in real time," he told New Scientist. "That's really an amazing thing that biology can do."
While this is certainly some amazing Twitch fodder, the team at Cortical Labs hope to use their findings to develop sophisticated technology using "live biological neurons integrated with traditional silicon computing," according to their website.
There's actually video of the brain cells playing Pong. The chief scientific officer told New Scientist that when the cells are in the game, they actually believe they are the paddle.
"We often refer to them as living in the Matrix."
The cells are placed on top of a microelectrode array that analyzes the neural activity... To teach the mini-brains the game, the team created a simplified version of "Pong" with no opponent. A signal is sent to either the right or left of the array to indicate where the ball is, and the neurons from the brain cells send signals back to move the paddle...
Brett Kagan, chief scientific officer at Cortical Labs and research lead of the project, said that while the mini-brains can't play the game as well as a human, they do learn faster than some AIs.
"The amazing aspect is how quickly it learns, in five minutes, in real time," he told New Scientist. "That's really an amazing thing that biology can do."
While this is certainly some amazing Twitch fodder, the team at Cortical Labs hope to use their findings to develop sophisticated technology using "live biological neurons integrated with traditional silicon computing," according to their website.
There's actually video of the brain cells playing Pong. The chief scientific officer told New Scientist that when the cells are in the game, they actually believe they are the paddle.
"We often refer to them as living in the Matrix."
do brain cells get tired? (Score:2, Interesting)
e.g. let's say researchers are finally able to integrate brain cells with silicon, would they be the weakest link, because they tire out, and need to rest far more than silicon chips would?
Re: (Score:3)
>would they be the weakest link, because they tire out, and need to rest far more than silicon chips would?
With a continual support system supplying nutrients it seems the best we can do now is about ~ 1 year before the organoids pass away :
https://www.spectrumnews.org/n... [spectrumnews.org]
Re: (Score:2)
Brain cells don't rest. When you're asleep the neurons are quite active and dream state seems to be vital to brain function.
Re: (Score:2)
Do they have to sleep?
Can anyone say Borg? (Score:2)
Because this is how it starts
Re: Can anyone say Borg? (Score:2)
They should call the cell culture Locutus.
On the other hand, if that blob of cells evolves into 7 of 9 I am all for it!
I am not convinced (Score:5, Insightful)
Re: (Score:3)
If there is not some sort of reward system, there really is no point for the brain to 'play the game of pong'.
Afaik, it just does something, there is no winning or losing, all is fine.
Re: (Score:2)
I am not convinced yet either.
If there is not some sort of reward system, there really is no point for the brain to 'play the game of pong'.
Afaik, it just does something, there is no winning or losing, all is fine.
This is astonishing: they have replicated the modern American educational system! Easy idle work, never stray too far from the norm, mediocre results, participation trophies for everyone!
Now if they can find a way to subject it to endless administrative bullshit and extract money from it, they can replicate the college system and get it a diploma so someone will hire it. Perhaps a Six Week Cybersecurity Boot Camp.
Re: (Score:2)
Re: (Score:2)
Re: (Score:1)
Re: (Score:2)
Nah, I don't think it needs a "reward" in the usual sense.
Of course it needs a "reward".
How else is it going to 'know' that it needs to actually hit the ball instead of, for instance, actively avoid it?
A child will often learn simply by moving objects back to a previous state, e.g. they pick up something because it fell.
A child is just as likely to stab itself with a sharp object or pull the trigger on your gun. Children don't do the right thing automatically and need a lot of coaxing, often by the use of rewards.
For example, if I spend a long time pulling weeds, then when I sleep my mind just keeps doing it over and over, simulating the effects and apparently pondering how to do it better.
You're confusing a small bunch of randomly structured neurons with an actual brain, which has a tremendous amount of genetically and environmentally determined structure.
Y
Re: (Score:2)
If there is not some sort of reward system, there really is no point for the brain to 'play the game of pong'.
It's hoping if it wins enough that the scientists will wire up a keyboard so it can communicate.
"K I L L M E K I L L M E . . . ."
Okay, I'm mildly creeped out about this experiment...
Re: (Score:3)
I am not convinced yet either.
If there is not some sort of reward system, there really is no point for the brain to 'play the game of pong'.
Afaik, it just does something, there is no winning or losing, all is fine.
I agree. Without an explicit reward, how does the brain know that the objective is to move the paddle toward the ball as opposed to avoid touching the ball? That the brain could figure out the unspoken objective would not be artificial intelligence but rather clairvoyance.
It would be perhaps interesting to learn the details of how this brain works. Unfortunately, the primary article is paywalled.
Re: (Score:1)
I was wondering about that too - the paddle covers about 2/5 of the space a ball could go through. It really didn't look any better than random to me, especially since the ball was often hitting the very edge of the paddle, with little indication of it actually tracking the ball, so seemingly only returning the ball by chance.
Perhaps there are better videos they could have posted that showed a more definite indication of it learning?
Re: (Score:2)
Agreed,
It looks more like a random walk than something responding to the environment.
Re:I am not convinced (Score:4, Insightful)
You don't even need to go that far. They're making some pretty outrageous claims, like " when the cells are in the game, they actually believe they are the paddle" which is pure fiction.
This is a grift, no question.
No chance of anything unethical about that then. (Score:2)
We can trust these people.
I'm sure.
Ethical Standards (Score:4, Insightful)
There is no standard ethics
There very much are ethical standards. Here [www.cma.ca] are the medical ethics standards for the Canadian Medical Association for example. Ethics may vary between individuals but there are definitely common standards against causing unnecessary harm and suffering to humans and animals.
Getting clusters of human brain cells to play pong raises serious questions about how aware those "mini-brains" are about what is going on and how they were motivated to learn pong e.g. was pain involved and, if so, how similar is this to e.g. torturing a severely mentally handicapped person? Experiments with human brains raise a lot of ethical concerns and having researchers goofing around by making them play pong at least suggests that they may be making light of those serious ethical responsibilities.
Re: (Score:2)
The clusters are about 1/171,000th of a human brain, in terms of the number of cells.
It's kind of hard to comparisons because they don't say how many are neurons, but assuming about half of them are (in humans it's about 86 billion neurons out of 170 billion total cells) then these brains would be similar to a lobster or fruit fly.
Kinda surprised about lobsters, TBH. I figured they were more intelligent than that, but surely Wikipedia wouldn't lie. Fruit fly seems more realistic.
Re: Ethical Standards (Score:2)
Experiments with human brains raise a lot of ethical concerns and having researchers goofing around by making them play pong at least suggests that they may be making light of those serious ethical responsibilities.
A Petri dish of brain cells.
Pong can be distilled to a one-bit input and output feedback loop, they didn't plug it into an Atari. Go worry over the fate of sperm cells in the shower drain, good grief. :%s/, dumbass//g
Re: (Score:2)
Pong can be distilled to a one-bit input and output feedback loop
Faulty reasoning like this suggests the gap in intelligence between a petri dish of brain cells and some humans is perhaps not so great.
A one-bit input is not enough for the three control options: move bat up, move bat down leave bat where it is. A one-bit input is not enough to say where the ball is and what its velocity is. This requires 4 pieces of information: ball (x,y) coordinates and velocity (x,y) components.
Re: (Score:2)
serious questions about how aware those "mini-brains" are about what is going on and how they were motivated to learn pong e.g. was pain involved and, if so, how similar is this to e.g. torturing
Nonsense. There are not 'mini-brains'. They lack practically all structure and all functionality of an actual brain. You are not born with a clean slate bunch of neurons in your head. The brain is actually highly organized on certain levels (and yes, the brain has many levels of abstractions it operates on).
Pain is a signal that is interpreted on a high level, not on an individual neuron level. Neurons don't have a concept of pain. They don't have concepts at all since concepts are functions of the neural s
Standardized ethics is the definition of law (Score:2)
> But, if something is coded into law, can it still be considered ethics?
Law is precisely the ethics used by a community, a society. Not just by some people in the society, but by the society itself, whatever structure it has.
Codified law is simply law organized into written sections that cover particular topics.
-
Then again, I *also* like a different definition of "law". My business law professor defines "the law" as " a prediction of how a court will decide a particular case". In societies with written
Re: (Score:2)
We can play the ethics are relative game, but ultimately I think it's easy to predict that a rather large population would be gravely concerned about experiments with living human brain cells.
Re: (Score:2)
I mean the game.
But really, that would be a good test, would pig brain cells perform better?
Re: No chance of anything unethical about that the (Score:2)
Hell this is a very interesting point. Imagine the uproar if the let shrimp braincells play against human braincells and the shrimp wins.
Re: (Score:2)
If they don't agree with YOUR ethics, then YOU should oppose this.
Bah, you only need about 100,000 human brain cells (Score:2)
Proteus, Demon Seed (Score:2)
Uh oh... (Score:3)
When they teach it to play Doom in 320x200, I'm beginning my search for a remote cabin in the woods.
Re: (Score:1)
We will find you...
Law of (computing) nature (Score:2)
Researchers at biotec startup Cortical Labs... (Score:2)
So.. (Score:2)
My parents were rightâ¦you need less than half a brain to play ping, eh?
Why not use neurons from other creatures? (Score:2)
Re: Why not use neurons from other creatures? (Score:2)
Maybe they can let the mouse cells play against human cellsâ¦
Petri Dish for President! (Score:2)
Brain under glass play computer games (Score:2)
Where's my fucking flying car?
Be concerned (Score:1)
Bio-neural gel packs (Score:2)
We may be reading about the early stages of yet another Star Trek concept [fandom.com] being brought from the world of science-fiction into the world of science-fact.