| If every token in LLM produces is statistically inferred (a fair claim) does that mean the entire output is a statistical inference?

Yes. Is the Pythagorean theorem an approximation ? Within the limits of measurement, yes. In geometric logic, no.

| Now what if I told you that a neuron in your brain can only be accurately modeled statistically? Is it statistical inference all the way up?
| Or do we accept the systems that are inherently based on statistical inference (broad definition, not narrow) can, to a high degree of accuracy, approximate
| logical inference?

Can only be accurately modeled statistically ? How long will you argue that this will be true ? Obviously, many disciplines (e.g., thermodynamics) make your argument, but that is only a feature of the model - it works better than anything we have found to date. Guess I'm a hard reductionist, but you may be right.

| Have you ever seen the results of a logic test described in a non-statistical way?

Yes! Multiple times in my high school geometry class, and thereafter -- many times to my sorrow. Also in sentential logic tests. The proofs/answers that I submitted were either correct or incorrect, and could be verified by deduction or induction.

| I fail to see how it isn't obvious that any kind of logical inference is merely an approximation.

You might take this up with George Boole, he is far more an authority than I. Snark not meant maliciously, I always learn from your commentary.

Do you have a counter-argument to the statement that statistical inference is not the same as logical inference ? I fail to see that an NN/LLM can perform logical inference. My understanding of NN/LLM proof systems is that they detect proof patterns in their corpus, but, having identified candidate proofs, submit them to Lean (or some other GOFAI, e.g., Prolog) for verification. Which suggests to me that NN/LLM cannot reason, which I define as predicate calculus. But I'd like to hear your view.

Should read 'Terafib'.

Waymo operates at SAE 5 ? Why does it need the Mechanical Turks and geofencing ?

Lean is a GOFAI symbolic logic engine. Combining NN LLMs with symbolic proof engines appears to me to be the way to go. NNs are statistical inference; Lean (and others) are logical inference.

Amen. I owned and rode a 1977 Ducati 900ss, an ancestor of Mr. Thompson's ride. It was bullet-fast and had the turning radius of an aircraft carrier. Every time I took it out for a frolic, I was dead straight stone sober. Every. Single. Time.

But at least Thompson did not do a Jackson Pollack pavement painting. My sincere apologies if I caused you any distress.

Sir: While Mr. Thompson's life choices were less than optimal, I defy you to find a better motorcycle road test than the Song of the Sausage Creature:

http://www.latexnet.org/~csmit...

And here is Cycle World's posthumous write-up of same:

https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.cycleworld.com%2F201...

My last remark was a bit of whimsy -- I would find it amusing if an unresolved great question in computational complexity was resolved by a computer. I wasn't trying to imply anything.

As for how long it's going to take, who knows ? Fermat's Last Theorem took 358 years, and lots of very bright bulbs took a swing at it, including Gauss and Euler (although both found a bit of traction).

So we will be having the solution to P ?= NP directly, then ? :) Sweet Jesus, maybe so...

I believe you may be confusing statistical inference with logical inference. LLMs use linear algebra to do the former. The latter is the province of propositional and predicate logic (expert systems, theorem provers, etc., for example, Prolog).

What constitutes a "thinking" LLM ?

Quantum blockchain - a distributed ledger where all transactions co-exist in superposition with their reversals. Now we're talkin' about some awesome accounting. Might have saved Elizabeth Holmes the prison time.

Given your handle, I would hope you remember Cover Oregon:

https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fen.wikipedia.org%2Fwiki%2F...

Should read "harebrained" instead of "hairbrained". Hares are probably more intelligent than hairs, but what do I know, I'm just chimp++.