You ever fuck an rtx 4090?

Could use chat to summarize it.

Absolute minimum for Sentience: capable of experiencing sensation/situation and deciding how to respond tto it.

This is arguably satisfied for any AI at or above the level of if-statements. Like even the ghosts in pacman meet this requirement. So that’s a bad definition lets step it up.

If you want to push the requirement up a single step you simply add: “capable of learning/unlearning responses to situations through experience (trial and error)” and voila all the “if statement” codes fail… as do all LLMs. In fact, any network model trained via gradient descent fails this definition.

Surprisingly (or not) some Reinforcement Learning nets don’t fail this definition though. But I’m not sure I’d call a spiking network that can classify MNIST images via STDP learning as “sentient.”

So, maybe we should push it a step farther and specifically say: “sentience is when something is capable of experiencing pain” And now all current “AI” fail.

Most general “Pain” meaning the unwanted interruption and disruption of other thought processes (network functions) by a specific stimulus which demands/steals the net’s attention/action/focus.

In order to pass this a network would have to not just learn and plan “in real time” so to speak, but also be capable of having those goal directed thoughts be interrupted/superseded by more salient stimuli.

Almost all (If not all) biological things with a central nervous system pass this. All current AI models fail.

This is where we could draw the line for “minimum sentience requirements” but even this is still rudimentary compared to what people typically mean when they say “sentience.”

Regardless, this is well beyond the capabilities of current AI. In fact, I’m reasonably certain that beating the ARC-AGI-3 contest (which can’t even be well attempted so far) with an actual learning-while-running model wouldn’t mean that net is capable of passing this definition of sentience. So this isn’t just beyond where AI is now but also beyond the major next milestone current AI has failed to reach.

Next Level (For Fun): Actual Sentience

The next step up would be the experience of suffering which is basically just the ability to recognize repeated/long-term/lasting pain; recognize that pain can/does disrupt thought processes; and experience the desire to remove/avoid the pain either due to it interrupting those desired thought processes. Again, this sentient system must still be capable of learning from mistakes which means capable of failing to fix the problem (end the suffering) and capable of realizing that it has failed. (If failure is also registered like pain, this gives rise to the possibility of creating a feedback loop of suffering / recognition of defeat/failure, congratulations you’ve created something capable of getting so frustrated by something it might be driven to invent things… or just driven to collapse into a manic/depressed non-functional mess… very relatable)

This, in my opinion, is a much more solid foundation for drawing the line of “this thing deserves at least some rights as a sentient thing.” Parrots and elephants and dolphins and dogs and cats etc. are likely at/above this level since they can experience frustration.

The next level up (for sentience specifically not cognitive function in general) would likely be the ability to predict/plan not just around the environment but around how the system knows it might react to possible scenarios and attempts to practice responding in a better way such that it will be easier when to respond better if/when those situations arise again/ever.

Any further levels would likely merge with consciousness requirements and other things like having a “theory of mind” or empathy. Further still might be the recognition of patterns in one’s own behavior that match those of others and give rise to the realization that one might have innate behaviors due to nature or nurture (specifications or experiences).

I would legit get a chuckle if it all just boils down to storage and recall.

You just just a sufficient amount of non volatile storage with redundancy and a high enough data throughput to use it.

But to even attempt to test that properly I think we might need to do some unethical bullshit.

But we do that anyways!

There’s a fruit fly playing Half Life Deathmatch

I suspect it would be since it would contain the same information, however, unless I was misinformed last time I looked into that (for scifi worldbuilding, so not terribly serious a search admittedly), the computational requirements to run a full physics simulation of a brain at that detailed of a resolution are probably just too high to ever be practical.

That being said, I really doubt you need to simulate one to the subatomic level, since the basic unit of a brain appears to be the nerve cells, and stuff more fundamental than that just provides the physical basis for those cells to work. Simulate the behavior of neurons accurately enough, and arrange them in the right way, and I’d personally be willing to bet theyd be sentient even if the simulation doesnt model every subatomic interaction within each neuron and just simulates something with the same output for a given input as an average neuron of the relevant type. The difficulty of that shouldn’t be understated, I would not be surprised if theres important but subtle behaviors in biological neurons that we’ve not discovered yet, and even if there arent, we’d want to be sure of it, and more importantly, “arranging them in the right way” is not (I would bet at least) a trivial problem, your brain is after all more detailed than just a lump of identical neurons that form connections at random, it has distinct structures that do distinct things, and that makes me think that the stuff we care about, like consciousness, is probably something resulting from arranging neurons in a specific kind of way, rather than just “thing that happens if you connect a large enough of them together”, else thered be no pressure for those structures to evolve.

On the other hand, I dont see any particular reason to assume that our brains are the most efficient or only way to do sentience. They exist because of a process of evolution acting on random mutation, not directed design, and while that can make some pretty efficient things, it wont find every solution, because every intermediate step has to be evolutionarily favorable as well and some things wont be that helpful half-made. As such, brains surely must represent at least an upper ceiling on the minimum resources required for sentience. A sentient thing using the amount of energy and matter that a human brain uses is definitely possible, because we observe them, but we have no indication that one smaller or more energy efficient cant exist. Maybe there are some processes that neurons do that have affects on their output, but which arent big enough to have an impact on sentience and can be abstracted away, for example. Im not sure exactly how that translates into the amount of compute you’d need, because brains are of course not digital computers, and simulating one with a digital computer is an extra step over just making a structure that directly does the stuff brains do.

Im not an neuroscientist or an AI or computer science expert and I imagine the intersection of those fields is where youd find the kind of person that could really give a useful answer. My gut feeling is that, given that humans dont need such massive amounts of energy to run or all the world’s information to train our minds, that our current way of doing AI is fundamentally extremely inefficient, and if some way can be found to create sentience with the technology we have, theres probably also a lot of room for making it more efficient too, to the point that I would be unsurprised if we have the compute needed already and just dont yet know how to actually use it to create that result. I kind of hope that its not figured out for awhile though- not because I think that a sentient AI would take over the world or something but just that, well, an actual sentient AI would hypothetically be a person, by many standards, and I see no way that, being created by corporations seeking a return on investment and with no legal rights whatsoever, it wouldnt just be a very high tech slave with all the ethical issues that poses and more. So, I hope that it takes long enough to figure that all out for the system that develops that technology to be replaced with something different, and hopefully more thoughtful for the ethics of it all. Thats probably wishful thinking on my part, but its something thats worried me about the whole AI boom. I dont think that what’s currently being sold is really as smart or capable as the marketing makes it out to be, but also that if it was, it wouldnt be ethical to use it as just a tool, and either way that doesnt paint the AI industry in a good light. Not that I really need to convince lemmy that theres something rotten there given its general sentiment towards AI.

I doesn’t seem like there is any good reason in physics that prevents this. The universe produces human brains and other animal brains here on earth a dime a dozen lol. I think we are nowhere near that yet tho, if we ever get there.