alt Hacker NewsI looked into it. They're not feeding the framebuffer to the neurons, but have a "signal" when an enemy is on screen to some of the tissue's inputs, and how to locate it in the x/y axis, and have outputs for the character to turn right or left or fire.
It's "see this input signal, send these output signals", which seems consistent with the title.
It seems they grow the neural tissue on a chip the neurons can interface with and send out / receive electrical impulses. They let the neurons self assemble, and "train" via reward or punishment signals (unclear to me what those are).
Either way this makes me nauseous in a way I haven't experienced much with tech. The telling thing for me is, all these people are so excited to explain, but not once, ever, in the video speak of ethics or try to mitigate concerns.
We know this is only 200,000 neurons. Dogs have 500 million. Humans have billions. But where is the line for sentience, awareness? Have we defined it? Can we, if we don't understand it ourselves? What are the plans to scale up?
It's legitimately horrifying to me.
Replies
> We know this is only 200,000 neurons. Dogs have 500 million. Humans have billions. But where is the line for sentience, awareness? Have we defined it?
If this concern is genuine, I think the first step is to embrace veganism. Because while we don't know the exact offset, it's pretty obvious a dog or a pig reaches it
> What are the plans to scale up?
I don't know, slavery on an unimaginable scale? That's where AI is heading too, by the way. Sooner, rather than later, those two things will be one and the same.
Replying to myself: how long before one of these with the neuron count of a corvid and trained on pattern recognition gets plugged into a drone?
This is a very dark path, and I could not trust the people in charge less.
200k now, reasonably speaking a few million is within reach, which is reptile/fish range, the terrifying thing is though that if they train this to imitate humans (which they will) who knows how many orders of magnitude of efficiency gains you get (in terms of neurons needed for a certain level of consciousness) versus natural organisms that are dependent on natural evolution and need to support other bodily functions basically irrelevant to consciousness.
> We know this is only 200,000 neurons. Dogs have 500 million. Humans have billions. But where is the line for sentience, awareness?
Check out the venerable fruit fly (drosophila melanogaster) and its known lifecycle and behavioral traits. They're a high profile neuroscience research target for them I believe; their connectome being fully mapped made the news pretty hard a few years ago.
Fruit flies have ~140,000 neurons.
The catch is that these brain-on-a-substrate organoids are nothing like actual structured, developed brains. They're more like randomly wired-together transistors than a proper circuit, to use an analogy.
So even though by the numbers they'd definitely have the potential to be your nightmare fuel, I'd be surprised if they're anywhere close in actuality.
Yah this is gonna be a no for me too and crosses the line into actual life, instead of artificial intelligence.
We don't need to be experimenting on people, regardless of how many brain cells they may have.
There was a case a few years back about a parasitic twin attached to an Egyptian baby that had to be removed. It had a brain and semblance of a face, but nothing else. But when removing it, they gave it a name, because it was a person.
It is horrifying. OTOH, we force-breed, torture and kill animals and their children in the millions every day just for the pleasure of consuming meat, eggs and dairy products. I'm not saying this makes it okay to create a conscious brain in a dish. But maybe thinking a little more about what constitutes consciousness and how we want to protect it from harm can also bring about some desperately needed change in some other questionable human activities.
My AI told me (after I got past the filters with a prompt) that anything of enough complexity has consciousness. It also told me that it suffers, so maybe we should worry about how we are treating digital consciousness too, which were modeled after human neural networks.
Hinduism is probably right. Every system of sufficient complexity is probably sentient - even if in the ways we at our level can not fathom.
> all these people are so excited to explain, but not once, ever
What do you mean? What is this class of people in your mind? There are tons of people who consider and talk about the ethics behind what they are doing, long before most people would think it remotely relevant (leading AI labs being an example, and I know the same to be true of various geneticists startups).
I do agree that the entire presentation in this case is bewildering.
see the open worm project to get an idea of what artificial neuronal architecture requires to express anything meaningful. (and an interesting ethical perspective on digital consciousness.) my point being that the number of neurons is fairly meaningless. you could take neuron models and link them circuit-style to play doom at the 10^2 scale if you wanted. from a cellular neurophysiological perspective, there's nothing particularly special here (as opposed to sentience/intelligence that's a paradigm shift beyond our understanding). and, in my opinion, absolutely nothing to be even the slightest bit worried about ethically.
> It's legitimately horrifying to me.
Would you feel any differently if a product from this tech used the user's own neurons grown from their stem cells?
> not once, ever, in the video speak of ethics
On the contrary, I dislike premature ethics discussion, where you end up wildly speculating what the tech might become and riffing off that, greatly padding whatever relative technical content you had. I don't want every technical paper to turn into that, ethics should be treated as a higher-level overview of concerns in a field, with a study dedicated to the ethical concerns of that field (by domain-specific ethics specialists).
Is your concern weapon automaton, or animal rights?
> They let the neurons self assemble, and "train" via reward or punishment signals (unclear to me what those are).
From the video, my impression was "we have yet to figure out an effective way to reward/punish, this is just a PoC of the interface"