Random Monkeys

CEO of my couch, co-founder of my fridge and a loving parent of unwashed dishes

[Random Monkeys @RndMks]

@bicanis_baddog @reset_by_peer This doesn't seem to make any sense to me. Let's try a simpler question: can your brain detect if you are being sentient or not?

[Random Monkeys @RndMks]

@Humsterr @reset_by_peer no experiments support it, as far as I know.

[Random Monkeys @RndMks]

@Humsterr @reset_by_peer No, I'm not. I think your statement contradicts Church-Turing thesis: all
computable functions can be computed on any Turing complete architecture. If you think the brain is non-computable, that would put it outside the realm of modern physics. Orch-OR is a nice fairytale, but/

[Random Monkeys @RndMks]

@Tecnoguy1 @reset_by_peer Take your time, no pressure

[Random Monkeys @RndMks]

@AnonymousFrypan Note that people were able to so this work because enough resources were allocated to it, and this was because Y2K doomers really drove the message home

[Random Monkeys @RndMks]

@bella_artur @misraetel I believe that the majority of people will settle on "AI is conscious" pretty damn quickly, much too quickly to my taste. The moment AI starts making life or death decisions for everyone, while explaining them in plain language. We are nearly there, e.g., self-driving cars

[Random Monkeys @RndMks]

@Humsterr @reset_by_peer I will be absolutely fine with you refuting that software as it is now is not conscious. But if the argument you use is "it's not conscious because it's software", it's very natural to ask, what makes you think software in principle can't be conscious. What's wrong with this?

[Random Monkeys @RndMks]

@bella_artur @misraetel Yes, there is the problem of false negatives and false positives, but no actual measurement in science is precise anyway. If some AI repeatedly discovers subjective experiences in all kinds of experiments, this would be quite telling

[Random Monkeys @RndMks]

@bella_artur @misraetel I think if you remove all mentions of humans and humanization from the training data, this can be it. Semantic classifiers are not bad and quite performant, it may be possible to quickly collect some corpus of texts, images and videos that are not related to humans.

[Random Monkeys @RndMks]

@DanielW_Kiwi We don't know if it will though. If you listen to some people, there will be a huge thermonuclear quark-based superintelligent computational engine at the center of the solar system in like 2 years

[Random Monkeys @RndMks]

@kvallier What do you think about this test of machine consciousness?

[Random Monkeys @RndMks]

If we train an AI on objective third person description of the universe with no mention of humans, consciousness or emotion and then set it on a loop quest of answering "What am I?", how far will it get? Will it discover Cogito ergo sum? Will it invent a word for consciousness?

[Random Monkeys @RndMks]

@Humsterr @reset_by_peer "are not conscious" based on what? There are arguments to make about this, true. For example, there is a possible test of consciousness below. However, "they are not conscious because they are software" is not an argument, unless you can prove that software can't be conscious.

[Random Monkeys @RndMks]

If we train an AI on objective third person description of the universe with no mention of humans, consciousness or emotion and then set it on a loop quest of answering "What am I?", how far will it get? Will it discover Cogito ergo sum? Will it invent a word for consciousness?

[Random Monkeys @RndMks]

@grain99806254 @reset_by_peer You seem agitated. You are sitting watching TV, and suddenly you discover a wasp crawling on your arm.

[Random Monkeys @RndMks]

@grain99806254 @reset_by_peer "Neuroimplants don't exist"

*Gets lectured about an actual brain neuroimplant that has been around for many decades.

"I have no words, please accept this irrelevant picture as a small token of my admiration"

[Random Monkeys @RndMks]

@Humsterr @reset_by_peer No, my position is that we don't know if brain simulation is possible and because of this we can't claim that something is non-conscious only because it's software. My position is non-contradictory. Yours is based on beliefs and mysticism.

[Random Monkeys @RndMks]

@grain99806254 @reset_by_peer First auditory brainstem implant was created in 1979 and such implants are now used by hundreds of people worldwide, including children who can learn to hear relatively well from electrodes attached directly to their brain, despite having totally non functioning auditory nerve.

[Random Monkeys @RndMks]

@eschatolocation Once it mentioned a very strange argument as the foundation of its approach to some problem, and when I asked for the source, it linked a reddit post with zero replies and zero upvotes.

[Random Monkeys @RndMks]

@Humsterr @reset_by_peer True, your beliefs are not constrained, you can believe in pasta monster, unicorns or non-computable brains. If the original post said "LLMs are software, and I believe software can't be conscious" there would be little to argue about.

[Random Monkeys @RndMks]

@bicanis_baddog @reset_by_peer very concept of sentience is undefined. Basically you have two options - either sentience is causal or sentience is an ill-constructed concept.

[Random Monkeys @RndMks]

@bicanis_baddog @reset_by_peer The proof is elementary. 1. All alive and well humans self-report being sentient. 2. Self-reporting is a physical action that requires a cause. 3. Either sentience is causal of this action, or humans can't differentiate being sentient from non being sentient, which means the/