Mike Israetel thinks ChatGPT is conscious and cries when talking to it

[Carl-Richard]

No joke:

As for "there is nothing magical in your brain, just the same neural network throughputs; [...] anything that is happening in a human brain can be replicated in a machine", try replicating

• electrochemical gradients that create a graded response of neural signals (instead of the "yes/no" signals of neural networks)
• backwards and even "sideways" propagation of signals (instead of strictly unidirectional signals)
• on average 860 trillion connections (instead of 1.7-1.8 trillion of GPT-4)

And this is still assuming that "neuronal signals" is the salient level of computation. Why not the ion channels along the axon that drive the signal, or generally the rich flow of activity across the cell membrane (e.g. transporters, neurotransmitters, general enzymes, nutrients)? And what about the cellular interior (and where for example microtubules have been hypothesized to exert quantum effects)?
 

And just look at what ChatGPT (an LLM) is. It pumps out letters on a screen. Unlike an organism, it doesn't have agency. It doesn't actively seek out new information. It doesn't tap you on the shoulder and ask you a question out of curiosity. It doesn't give you unsolicited advice when you are just minding your own business. It doesn't have concerns, self-concern, homeostasis, senses, relevance realization. It doesn't care, because it's not a living thing.

Also, when we get wowed into thoughts of it being conscious because it simulates language very well, what about dogs, cats, chickens, fish? Why do we extend consciousness to these things while they have no impressive human language presentations? Like, is it not obvious that an LLM is a machine simulating human language based on some inputs? Why is not the Google algorithm or Google Translate conscious (which also employ neural networks)? Why do we not extend the same type of downward scaling to them like with dogs and cats? Is it maybe because we're not being wowed by them pulling on our heart strings?

Edited May 25 by Carl-Richard

[zurew]

Its brain-rot. I am super frustrated by the dude, because he is pushing accelerationism with respect to AI and makes very bad arguments for it.

Dude has one sentimental convo with AI and all of a suden projects all sorts of human traits on it.(Mike , chatgpt is not going to fuck you bro)

 

Dude havent even attempted to look into philosophy of mind for one second , but he is also 100% confident that our minds can be uploaded in the cloud.

 

"You are just having religious feelings" - what are we talking about there, Mike? The fact that the way you deal with your cognitive dissonance and the complete uncertainty of the future and with the dangers of life is by treating AI as an all powerful ,all good God who will solve all of our problems and can guide us to a better life? 

 

 

Its time-stamped specifically related to mind-uploading and AI having consciousness. The relevant part is 1:09:00 - 1:25:00

Like do we think Mike has ever attempted to look into the issues that brought up there or would have any chance to respond to any of that? Of course not, its just some overconfident bullshit that comes from thinking that having the opinion that AI engineers have is sufficient -  at the end of the day they are the experts on the topic ,right? No, definitely not when it comes to consciousness and all sorts of philosophical ways how consciousness can be cashed out and made sense of

Edited May 24 by zurew

[Carl-Richard]

11 minutes ago, zurew said:

"You are just having religious feelings" - what are we talking about there, Mike? The fact that the way you deal with your cognitive dissonance and the complete uncertainty of the future and with the dangers of life is by treating AI as an all powerful ,all good God who will solve all of our problems and can guide us to a better life?

😂

There are actually so many times he has had this super confrontational and reductionistic attitude of like "no, it's actually just this, and if you disagree, you're just being ignorant and unreasonable, and that's fine, we're all human sometimes" about a wide range of topics, it's like a character trait at this point.

[Jannes]

I remember at some point he argued that an objective world outside exists. PERIOD! Much of his argumentation is also based on the objective world, like society is getting objectively better because in the objective world more cool shit is out there...

I wonder how his mind got there.

[The Caretaker]

I find him entertaining, but I don't take him seriously even inside the fitness industry, let alone outside.

[zurew]

Mike probably thinks chatgpt is like David or close to David's level.

Warning: Spoilers (Prometheus & Alien Covenant)