Self-evidencing of a scientific observer O bears no explanatory content except proving that science that presupposes O will never explain O.
A successful explanation of a scientific observer O describes our universe prior to O in a way that explains O first, then how it appears to O.😊
So now we have automated the Punch and Judy puppet show version of a consciousness conference.
Yeah let's burn the biosphere with that.
😐
You need to read this if you want to understand how complexity works in brains.
I think this might be a pivot moment.
Finally somebody has put down how my mental vision of it works.
The brain is Turing Computable iff you already know everything! Crazy energy commit by computers is explained.
However it turns out, this article by Todd is a pivot point in complex systems applied to brains. So pleased!
"Artificial neural networks ...."
Drives me crazy. They are not artificial in any scientific sense, but merely as "made by humans".
Neurons are natural objects!
An artificial version, in the science sense, is like an artificial kidney is to the original nature.
So tired of seeing this mistake.
You bet. 😊. It doesn't even begin to 'explain'.
The explanatory emptiness of correlates is a scientific idea cemetery because it presupposes the consciousness of US. The scientific observer.
Until we accept this tediously obvious context the science is dead.
doi.org/10.3389/fnhu...
Those moments are treasures of experience that the nagging of our tech does nothing but diminish. The photo is in your mind and conveyed brilliantly by your description. All we need! That's the magic of consciousness and associative memory.
Because AI does not work like the brain.
www.nytimes.com/2026/01/26/t...
#neuroscience
I had trouble answering your question. Excitable cell EM field theory is in a confused state. A 1 page summary for discussion is below. It answers your question. Happy to chat about it if you have time.
@gauteeinevoll.bsky.social too. If you want.
This area is slowly coming into focus. At last!
It's the population of 50000 that collectively expresses all the spectral content found in the travelling EM field activity. Alpha, Beta, Gamma ... All of it. It all comes from the EM field of the membrane-embedded action potential sources.
It's the 1st thing we throw out in neural models!😊
I have been trying and failing for 15 years to get this point made.
I have literally built an inorganic piece of membrane and "shorted it out" (ion channel). Guess what? It makes the same kind of EM field. Lipid bilayer has dielectric constant of mylar. Works fine!
The video!
If you delete spikes there would be no LFP! Spikes literally make the LFP. Here's 1 spike making its LFP contribution. 50000 nearby neurons, resonant, making a travelling wave of EM field, literally superpose into LFP. The rotating dipole is of the electrical potential created by the field.
These deployments are both doomed in one sense: failure to capture the human expert's capability to autonomously acquire new knowledge in the target domain. Knowing what is not known, finding and integrating it, creates the real authoritative value of domain expertise. Both examples fail this.
10 years+ since it came out.
Kuhn warned me about paradigm shift disorders. We are still in it.
www.worldscientific.com/worldscibook...
It's 2025 dusk, more than 10 years since my book on the science of consciousness and the making of an artificial scientist.
I had a look for it. Yep. There it is.
The preamble is free. I read it. Damn. It has withstood time! Go me!
www.worldscientific.com/doi/suppl/10...
Congrats!
But I bet there's nothing in the work about "what it is like to be" a member of a set. Or perhaps "what if I act like a member of set X?" is another perspective. Nobody ever formalises this and then finds the physics that does it.
I've been looking for >20 years for it. I live in hope!
There's a lot on "scientific understanding" in my book on building an artificial scientist.
If you want the PDF email me at unimelb.edu.au.
www.worldscientific.com/worldscibook...
It's time to have this issue out.
This long lost era of mistaking the computation done by natural signalling for a computed abstract model of properties of that signalling ... Has to end.
That the latter is/can be an artificial version of the former Is an unproved hypothesis at best.
Spot the neuron membrane in the photo?
No. What about experimental replication of the exact physics that is brain signalling? Observing it as a dynamical system is an untried option. No model. No computer.
Here it is. Normal science untried. Until now. Untried because physics and neuroscience haven't met properly yet.
Modern AI is neither A nor I .... Because actual artificial versions of natural brain signalling (not a model) have never been built.
Actual AI is in the empty spot in the diagram. An artificial version of a natural thing NEVER happens in the middle. That is can in this 1 place is a hypothesis.
But you do not know what you are not doing. The brain that does function X is easy to model:you defined what you look for. In that definition everything not in the model is lost.
What change/learning is being missed? That's what this is about.
The real empiricism part does not involve models.
It never ceases to amaze me that the one thing never tried remains untried and invisible when you are sitting behind a computer ...😊...
Making an inorganic artificial version of natural brain signalling physics. No computer at all. Analog or digital.
Spot the neuron membrane in this photo?
Paragraph on the missing empirical technique (what I am working on in my lab)
Described in great detail here. In particular the missing in silicon empirical method that normalises the science (bitmap in sub post).
www.frontiersin.org/journals/hum...
Bitmap of the important para.
Physics/neuroscience will settle it in the end.
If the LLM has no phenomenal states, then it has no 1st person perspective. To "be" the LLM is to be what we experience as a coma.
That being the case, "self awareness" will be whatever the LLM appears to say of itself, to an observer of it.
The word "awareness" doesn't seem apt!
