Scott Aaronson – The Search for Physical Correlates of Consciousness

6 thoughts on “Scott Aaronson – The Search for Physical Correlates of Consciousness”

1. If he know something about consciousness he would act differently, if u want know something about consciousness chech out Srila Prabhupada

2. I would like to point anyone here to Tononi's response to this argument (unfortunately, it is not well represented in this video):
   https://www.scottaaronson.com/tononi.docx

   It should take about 20 minutes to read (like the time needed to watch the video), and is an interesting text for anyone interested in consciousness, why we need a theory, what IIT says, and why Aaronson's argument is misguided. Enjoy!

   edit: here is also a link to the response to the response for anyone interested:
   https://www.scottaaronson.com/blog/?p=1823

3. He should take his own advice: start with the clear cut cases before moving on to the more difficult ones. That means start with tests in normal waking humans capable of trustworthy reports, then move on to other normal states (sleep with and without dreams or anesthesia etc). Check whether the theory agrees with those cases, and other evidence about the brain. This has been done for 20+ years now with IIT, and it has been quite successful.

   Then, much much later, you may start considering generalizing to more complex cases where reports are not available (eg animals, machines, organoids etc). You may even speculate about unconfirmable thought experiments like this one. But before mudding the waters in a field you are not trained in, have the decency to grapple with (or try to understand) at least the most basic concepts and notions of the theory you are throwing under the bus. This is a weak example of uneducated guesswork and flaunting of untestable intuitions.

4. The crux of Aaronson's crticism is that his NAND gate grid is not concious (as Tononi suggests), and therefore IIT is falsified. But how on earth can Aaronson make that assumption? The grid could have a form of consciousness which is completely different from human consciousness. Aaronson seems to be falling into the trap in thinking that all conscious systems must resemble a brain. Aaronson compares his supposedly unconscious NAND grid to a wall – but they're completely different. The NAND grid is processing information in a complex manner whereas the wall is not.
   In contrast, Tononi's statement that the cerebrum being conscious and the cerebellum being unconscious is evidience for IIT stands up because there is clear evidence from neuroscience that that is the case (using introspection and brain scanning). This is, indeed, evidence that IIT is correct.
   The only way to falsify IIT would be to find a part of the brain which is unconscious when IIT predicts it is conscious, or vice versa. Aaronson's NAND gate grid does not falsify IIT as we have no idea if it is conscious or not.

5. So, Scott starts out by explaining it is impossible to measure conciousness and then goes on to talk about the "obvious cases". But since we can't measure them, we know nothing about the conciousness of these "obvious cases". In other words, he is making an assumption with, as he himself states, no evidence.

   Sure, if that assumption holds true, then IIT wouldn't make sense. BUT, it is pure conjecture to assume that the Grid of XOR gates has to be less concious than a human. He doesn't know. It's quite arrogant to assume that Humans must have the highest amount of conciousness. It is equally arrogant to assume that electric circuits can't be concious.

   This is a bad attempt at disproving IIT. By making random assumptions, you can disprove anything.

   Hint: Never make random assumptions.

6. Φ is an emergent property of "a bunch of verbal discussions." Priceless.
   I often wonder if the placeholder descriptor consciousness has a detrimental effect on members of our species who are engaging in exploring the natural world.

Comments are closed.