That's the thing with metaphors, they have similarities but there are also points of divergence. Point is, my metaphor was not meant to be understood as a technical description of the system's workings.

For the unsuspecting soul who approaches this modern oracle without the faintest idea of how it works, the experience of facing unexpected demons could serve as a warning of the dangers they may face if they approach the tool without caution.

People compare AI and robots with Frankenstein's monster (or with Pinocchio, on a good day, if they want to give the story a positive spin), the construct which gains a life of its own.

But current LLM chats are more aptly compared with a ouija board. The machine itself is inert, and you can see it as a playful activity. But the model contains within it the highlights of a whole culture compressed during its training. You can access the souls of all the authors whose works were used for learning; but also of all the internet fanatics, trolls and scammers. When you set the machine in motion, you never know whose spirit are you invoking to answer.

s/how the person mental started/ how the person feels mental states/

There is no room for it to manifest in a computer program. There is no room for any "magic" in computer programs.

That's true for classic software in a trivial way, in the sense that a sequence of logical inference steps (i.e. a deterministic symbolic program) do not reflect upon itself.

However it may be possible that the computer program is not conscious, but the computer running the software is. LLMs in particular generate their output not from the specific instructions included in the program, but from the weights trained in the model; the software instructions are a requirement for the weights being interpreted, but the outcome doesn't necessarily follow the rules of a formal system and an inference process.

Current LLMs do not have consciousness because their processing is too simple for it to emerge; not because the software substrate is deterministic and mathematical. If the base software were processing the weights of the model in ways similar to how neurons generate brain waves, it is plausible that the emergent system-level information patterns appearing at the data level could exhibits the attributes of consciousness, including self-perception and self-reflection. This is true even if the computer software is deterministic, in the same that the neurons in our brain behave in deterministic electro-chemical ways.

On the contrary, it means that neuroscientists have measured precise ways in which brain waves of vision and audio processes converge into taking decisions before the person reports being conscious of taking such decision; and that they have studied precise ways in which altering the brain chemistry affects how the person mental started. Just look for the papers on these experiments for these topics.

No way. We may not have a full scientific understanding, but neuroscience has made huge advances in how consciousness emerges in the brain and how it is affected by the changing conditions of its low-level processes.

We cannot say that machines at some point will never have similar emergent patterns that could become conscious. But we for sure can say that the current ramblings of text generation from LLMs definitely can't be conscious, because they are created directly by much simpler low-level deterministic computations.

The long LLM-generated dissertations that people mistake for conscious reflections do not come anywhere near from the complex introspective processes that we know are involved in having consciousness; they are just mechanic pattern generation from the highly compressed encoding of human culture one which they have been trained. It's true that our own brains do learn by highly compresssing our live experiences, but we know for sure that our consciousness involves something more than just compiling memories.

That nails it. Too many people think that AI models are either Pinocchio or Frankenstein, a constructed being who gained a life of its own, becoming friendly or terrifying; when in fact the current batch is nothing more than The Wizard of Oz, faking the appearance of an awesome entity because some human behind the curtain benefits from making you believe that.

If it can, then it breaks the deterministic behavior of the known and understood physical components.

What makes you believe that? Our current best understanding of consciousness is that it's an after-the-fact rationalisation of the multiple low-level brain processes that converge into a subconscious decision. If that's the case, consciousness doesn't influence the external world in a non -deterministic way.

If LLMs are not conscious it's because they don't have this high-level aggregate feedback loop, not because consciousness needs to be non-deterministic. All their outputs are created from low-level reactions, like the reflexes of an amoeba that grows in its environment towards the gradient with more food.

Paper and glass boxes is the best technology for voting. It's easy to spoof, but it's also easy to detect when someone is spoofing the result.

The most essential property of a voting process is that anyone can understand how it proceeds, and with digital voting you need to rely on techno priests reviewing that everything went as intended. That's not secure, no matter how complex cryptographic algorithm you create to avoid tampering.

Because laws must reflect whatever society collectively considers ethical. The monopoly is granted for an ethical reason (at least in theory), so corporations must comply and accept that, even if they like to pretend it was created only so they can extract profit and ethics was not concerned.

If you believe that computers are fully deterministic, I have a PC I want to sell you...

Theoretical computers may be fully deterministic, but physical computational machines are made of electrical signals running on rare earth semiconductors, and with AI we have complex statistical chaotic interactions on top.

Any small unpredictable perturbation at any later may swing the whole system in a whole new direction. Hardly what we'd call deterministic (unless you believe the whole universe is deterministic, in which case the word loses all differentiation power).

So, basically the weights of an AI model are a compilation of human knowledge, i.e. an encyclopedia - or its equivalent for the XXI century. They don't contain just what human authors select and have the capacity to write, but everything that they're exposed to and that is frequent enough to leave a trace in the weights.

Any content generation of an LLM which isn't guided by a human is merely a random walk through that compressed encyclopedia of everything, following the most well-connected paths more frequently (that's why long generations often get caught in loops btw). There's no real agency in those 'agents'.

None of them. They're borrowing trillions in a triangular scheme of circular lending; the computing platform company invests trillions in the AI company which invests trillions in the energy company which invests trillions in the original computing company, increasing their overall valuation until they're all Too Big To Fail before the bubble pops.

+1 Informative, you

If a new development is going to leave large numbers of people unemployed, it would make sense for the government to seize it so that the wealth created by it can be invested in supporting those who lost their jobs because of it.

However, for some reason I suspect Anthropic's CEO may be hoping for the opposite result, having the government subsidizing the private company with whatever excuse.