Convincing AI: A New Approach to Settling Scientific Debates
Samir Varma joins a dialogue on free will.
I just received a publicist's email about a new book, The Science of Free Will, by Samir Varma. While Dr. Varma looks like a decent guy with proper credentials in physics and other "sciency" stuff, I was ready to swipe. After all, I've covered this debate a bunch of times and many of the leading minds in physics, neuroscience, and consciousness studies have long ago moved past the "you're a biological robot in a meaningless universe" narrative.
But beyond being a legit physicist with stellar credentials, Samir is also well-versed in AI. Aha—this could be an opportunity to dig deeper into the AI Truth use case I've been hammering on. Especially since daily advancements in AI reasoning/thinking are dominating the headlines and possibly hurling us toward a future where machines, not humans, will become the most trustworthy voice in scientific debates.
I invited Samir to join me in an AI-mediated dialogue, and he agreed.
Let's Go!
I'll kick things off by recapping the traditional physicalist argument:
We're made of atoms. Atoms follow deterministic laws. Therefore, free will is an illusion.
As scientific dogma goes, this is standard stuff. Unfortunately for Samir, this dogma sidesteps mountains of recent lines of evidence that contradict it. Many of these findings fall under the umbrella of "mind-matter interactions," and I've chosen this as the starting point for the dialogues with Samir.
Next, I had to consider how to bring AI in. For this, I found inspiration in Samir's response to my suggestion that mind-matter interactions challenge his claims. He flatly stated, "I'm unconvinced." Brilliant! This perfectly frames up the problem. I maintain that humans, burdened by our never-ending biases and limited understanding of logic, should not demand to be convinced by AI. Quite the opposite. In the very near future—like 30 minutes from now, when I pose this problem to AI—it will be humans who must convince AI.
One Prompt
I began by loading Samir's book into a Notebook LM, which generously accepts long-form books. And then, with just a single prompt, I got this:
My Prompt: Wouldn't scientific experiments demonstrating mind-matter interactions contradict the physicalism and determinism? After all, mind-matter interactions are, by definition, instances where consciousness affects matter. This directly contradicts Samir's proclamation that consciousness is governed by deterministic physical laws. I don't see how he can resolve this contradiction - do you?
Notebook LM: Your query raises a critical point about the implications of mind-matter interactions for the author's claims about consciousness and determinism. You're asking if such interactions would fundamentally contradict the idea that consciousness is governed by deterministic physical laws. Here's a breakdown of why this is a challenge for the author's framework, drawing upon the sources, our conversation history, and some logical analysis:
• The Core of the Author's Argument: The author argues that everything, including consciousness, arises from deterministic physical processes. He posits that the universe is governed by a set of rules, and all outcomes are a result of following those rules. He differentiates between Free Will in Theory (FWIT), which he says does not exist, and Free Will in Practice (FWIP), which arises from computational irreducibility, i.e., the impossibility of predicting the outcome of a system without running the system.
• Mind-Matter Interaction as a Challenge: You're right to point out that mind-matter interaction is, by definition, a case of consciousness influencing matter. If the mind or consciousness can directly alter physical states without any physical intermediary, it poses a significant problem for the author's deterministic and physicalist framework. This is because the author defines physicalism as the idea that everything can be explained using known laws of physics. If the mind can act on matter without being governed by those physical laws, this would violate the author's view that all of biology and human behavior is entirely contained within known laws of physics.
• Deterministic Laws and Causality: The author claims that, given a complete physical state at a time T1, one can exactly determine the physical state at a later time T2 via physical laws13. A mind-matter interaction, especially one that is not predictable from previous physical states, would violate this deterministic causality14. If consciousness, something that the author claims is an emergent phenomenon arising from deterministic laws, could circumvent the deterministic laws that are said to define it, then the entire foundation of his argument would be undermined. It's not a matter of some unknown level of complexity, but a matter of the mind influencing the physical realm without any physical basis for such an action.
• Computational Irreducibility is Not an Escape: While the author emphasizes computational irreducibility as the reason we can’t predict human actions, this concept is still within the bounds of deterministic processes. Computational irreducibility means that there is no shortcut to determining the outcome without running the process. If mind-matter interaction bypasses the physical laws, even in an irreducible way, it means that it is not simply that we can't predict the outcome; it means that the physical laws themselves are not determining the outcome, and this breaks the author's model of the universe. It would also mean that the mind would not be an emergent phenomenon but rather a separate force that can act upon the universe in a way that is not determined by physical laws.
• The Problem of Non-Physical Influence: The author's model is built on the idea that all interactions are physical, i.e., they follow deterministic laws of physics2.... If mind-matter interactions involve some non-physical influence on the physical world, this would contradict the very basis of his physicalism. The author explicitly denies the existence of nonphysical souls, life forces, or creators, but if the mind is shown to influence the physical, without physical intermediaries, it would be doing exactly what these non-physical entities would do.
• The Author's Stance on Non-Determinism: The author argues there is no non-determinism at the fundamental level of reality, and that claims of indeterminacy in physics are either inconsistent with science or rely on faith19. A mind-matter interaction that bypasses deterministic physics would require that the mind (or consciousness) has its own set of laws, rules, or forces, independent of those of physics, and the author does not accept such a possibility.
Where the Author Might (Try to) Push Back
• Complexity as an Explanation: The author might try to explain mind-matter interactions as a result of incredibly complex interactions within the brain3..., arguing that we just don't understand the full extent of this complexity, and that it is still governed by deterministic laws3. However, if the interaction can be shown to involve a non-physical influence on physical matter, that explanation would not be sufficient. He would be obligated to demonstrate a physical basis for this action in order to retain his argument.
• Redefining "Matter": It is possible the author would try to redefine what "matter" is, and perhaps try to include some aspects of mind within that definition. However, that strategy seems unlikely given that the author is an avowed physicalist.
Why This is a Problem for the Author
The problem here isn't that the author's model doesn't account for human unpredictability. He addresses that with computational irreducibility and FWIP. The problem is that mind-matter interaction implies that consciousness has a causal influence on matter which violates the laws of physics without any apparent physical cause. This would undermine his core claim that consciousness is entirely governed by deterministic physical laws. It wouldn't simply be a case of our inability to predict behavior; it would mean that deterministic laws of physics are not the only things governing the universe. This would be a huge problem for the author's core thesis.
In conclusion, you are correct that mind-matter interactions present a major challenge to the author’s argument that consciousness is governed by deterministic physical laws. They would constitute a violation of the physicalism and determinism he espouses, and it's difficult to see how he could logically reconcile such a contradiction.
Let's see if Samir can convince AI to change its position.
#TruthSeeking #Skeptiko
It's becoming clear that with all the brain and consciousness theories out there, the proof will be in the pudding. By this I mean, can any particular theory be used to create a human adult level conscious machine. My bet is on the late Gerald Edelman's Extended Theory of Neuronal Group Selection. The lead group in robotics based on this theory is the Neurorobotics Lab at UC at Irvine. Dr. Edelman distinguished between primary consciousness, which came first in evolution, and that humans share with other conscious animals, and higher order consciousness, which came to only humans with the acquisition of language. A machine with only primary consciousness will probably have to come first.
What I find special about the TNGS is the Darwin series of automata created at the Neurosciences Institute by Dr. Edelman and his colleagues in the 1990's and 2000's. These machines perform in the real world, not in a restricted simulated world, and display convincing physical behavior indicative of higher psychological functions necessary for consciousness, such as perceptual categorization, memory, and learning. They are based on realistic models of the parts of the biological brain that the theory claims subserve these functions. The extended TNGS allows for the emergence of consciousness based only on further evolutionary development of the brain areas responsible for these functions, in a parsimonious way. No other research I've encountered is anywhere near as convincing.
I post because on almost every video and article about the brain and consciousness that I encounter, the attitude seems to be that we still know next to nothing about how the brain and consciousness work; that there's lots of data but no unifying theory. I believe the extended TNGS is that theory. My motivation is to keep that theory in front of the public. And obviously, I consider it the route to a truly conscious machine, primary and higher-order.
My advice to people who want to create a conscious machine is to seriously ground themselves in the extended TNGS and the Darwin automata first, and proceed from there, by applying to Jeff Krichmar's lab at UC Irvine, possibly. Dr. Edelman's roadmap to a conscious machine is at https://arxiv.org/abs/2105.10461