32 Comments

Reminds me of discussions with my husband.

Expand full comment

Clever!!! :-)

Expand full comment

Merry Christmas:

This is logic theory in a binary world. But in a quantum physics of nonlocality, superposition, can't logic rules be rewritten? This is what makes us so human? We are nonlocal quantum supercomputers, as with the Orch-OR model of consciousness portrayed by Dr. Stuart Hameroff and Sir Roger Penrose. So if so, how does this affect our concept of logic?

https://hameroff.arizona.edu/

Expand full comment

Seems like our systems could instead be trinary, 0, 1 and 2, that would begin to approximate reality better. Or just abandoning the 0/1 dichotomy would be helpful, and not reinventing it anyway. Paper and pencil again. Best

Expand full comment

Thanks, Jacquelyn. However, it seems to me your proposed trinary system would be a quantitative, not qualitative improvement.

We're back to von Mises dictum about why communism failed.

Expand full comment

I mean within the mind of the computer, more than binary. Not in terms of class.

Expand full comment

Excellent post. Thank you.

And best for Christmas to you too... .tho in the binary world of AI, one would never proffer a blessing (whatever THAT is to AI) to an unknown entity on Substack, wishing good (whatever that is; in fact, in a amoral - n.b., not immoral - world, good also has no definition, exactly as CD Lewis noted in the Abolition of Man) in an unknown amount.

Expand full comment

Merry Christmas, Dr. Jack!

Your brain is scary good...

Expand full comment

Interesting; has anyone accused you of having too much free time? Not likely.

Merry Christmas!

Expand full comment

LOL no, just not enough sleep!

Expand full comment

Yep, it's broke. Thanks for (your extreme patience in conducting) this experiment to explain why. I wrote out a small thing (AND, OR, and/or AND+OR?) on this about a year ago. Your experiment made me remember it, so here goes.

Life seems about separating the 'OR' propositions from the 'AND' propositions. For example, AND examples include the cumulative effects of layers or lacquer, of dumping trash, of head traumas or air pollution. The OR propositions, where just one path will be chosen, when '2 paths diverge in the wood' (less so in the 'hood', the street grid allows a return to the same spot much more easily than in the wood). OR propositions include whether you passed a test or not, the machine works, or not. Zero OR 1, not both. In a digital world of 0 or 1, there cannot be any true AND statements, all must be OR. This is where we find ourselves, having adopted digital systems that essentially narrow our choices as we interact with it. Digital systems are a terrible model of reality as they essentially constrain it to binary, and reality is not binary. I am glad I did not have to talk to an AI machine to discover this, I think I was pondering the kinds of mistakes the Gates mentality would bring us when it is allowed to influence all we do. It's quite bad as we can see, reductionism kills.

Expand full comment

False dichotomies are treachery - and nature, being complex, abhors the imposition of the expectation of parsimony!

Expand full comment

If we think this is a fun way to spend Christmas, what does that say about us?

😁

Expand full comment

It says you're a Popular Rationalist! :)

Expand full comment

oh my gosh I can't believe I read this whole thing! Captivating actually, although I did lose you a bit along the way. How geeks spend Christmas! :)

Expand full comment

Awesome, thank you! Merry Christmas Zora!

Expand full comment

You are usually up to something interesting. Thanks for sharing. Thank you for being a good human being. I appreciate you.

Expand full comment

Ok...firstly, your AI gives me shivers up my spine....probably some leftover scars from dealing with narcissists.

Second, it definately is lacking in its truth table capacity.

And thirdly, on "absent empiricism, one can say anything - but merely saying something does not make it true"

I'll leave this right here: "62,400 repetitions make one truth" A. Huxley

Do you think the psychotic brain operates with the same set of rules of logic and sense of reality as we do, or is it more akin to trying to rationalize an animal's thinking, or an aliens?

Expand full comment

1. I agree, AND vs OR is an essential detail to which one must attend. So are quantifiers.

2. I’m not sure I understand your point. Your boat example is flawed if you’re trying to show a conflict between formal logic and reality. All you’ve really illustrated is that AND doesn’t match that context. But anyone would have predicted that by way of your definition of TRUE := sink in that setup.

3. Didn’t you use the ANDness of “contradiction” to illustrate that the chat bot made contradictory statements? I don’t understand what you’re trying to illustrate.

4. You can have more than binary logic. You can even define it to take values on the continuum! Not a lack of creativity, but a tool that you use based on the situation. There’s also second order logic, fuzzy logic, etc. lots of choices if you don’t like binary. But adding “contradiction” as a label could only apply to a composition of propositions, so by itself it can’t be a label at the same level as the labels TRUE and FALSE, which can be applied to primitive propositions and compositions alike.

5. I think your un-careful treatment of quantifiers is making it hard to fully understand your point(s).

6. Gödel proved that any system sophisticated enough to encode the integers is incomplete. He also showed that the language of a model can’t be used to prove things about that model. Seeing as he used binary logic for all this, I’m not sure we can extend his result to binary logic. You’d need some tool that isn’t binary logic to prove binary logic is incomplete.

7. There’s a wide gap between reality and the philosophical vacuum (that’s my name for it). Certainty can only be had in the philosophical vacuum, and cannot be had in reality. Therefore, the truth table for reality/empiricism is easy. UNCERTAIN. If you allow the binary system of logic to be employed in the context of reality, the best you can attain is a truth table that includes FALSE. TRUE can never be had because empiricism can only process falsifiable statements. In short, the words TRUE and FALSE are abstract and aren’t empirical.

8. As to the question of whether the statements “All X are equal and no X are equal” and “All fish are birds” are false in the same sense… well, we’re gonna need careful definitions to answer that. If we take the naïve definitions, then yes they are both false in the same sense. They both violate those naïve definitions of the words in the statements. That’s what makes them both false. They are both contradictions in the philosophical vacuum only. If you mean that the latter can be demonstrated by holding a fish in one hand and a bird in the other, then I will agree, because of what we mean by bird, fish, and is/are. The exact same thing is happening in the other statement, and not being able to behold those things manifest doesn’t mean it’s false in a different sense. Because the designations of true and false are abstract. Unless you want to redefine the labels or words. Ok. That’s fine. And if those definitions prove useful for awhile, people will use them… until yet better definitions are formed.

9. I’m not sure you’re undermining perceptions of rigor, but perceptions of absoluteness.

10. I think you’ve done more to persuade us that the chatbot is closer to human than we thought. Because humans make those errors of inconsistencies all the time and don’t notice, and then try to weasel out of it once confronted. I jest, of course. I think you’ve effectively demonstrated the opposite. Unlike humans it has no ego. It’s merely a sophisticated lookup tool.

Expand full comment

Did you see Steve Kirsch's interview with a young medical student debating whether vaccines are safe and effective. The student asserts that vaccines are safe because he knows of no mechanism that could cause lack of safety. Steve argues that vaccines are unsafe because all cause mortality has increased by over 20% since vaccines were introduced. Insurance companies report large increases in younger populations not expected to die. The student stated he could not evaluate statistical arguments since he didn't know statistics. He also stated that all cause mortality was not a useful measurement since vaccines should only be measured against Covid deaths when proving their efficacy. Other causes of death are not relevant for vaccines.

He did agree that injecting a bolus of spike protein directly into the blood stream through failure to aspirate might be a cause of death--but only happens in about 1 or 2% of injections. (NOTE: My LabCorp phlebotomist stated they do not aspirate when vaccinating because they can look at the skin and see the veins when they see blue color. Also there are no veins in the spot where they inject.)

The student didn't accept any deaths immediately after vaccination as being caused by the vaccine, because he knew of no mechanisms that would account for such deaths. The student seemed to be genuinely trying to consider Steve's arguments. Several times he repeated an argument and asked Steve if he had understood him correctly.

Steve gave the student a paper to read that discussed vaccine injuries. One author was Schwab. They agreed to meet again for further discussion.

Expand full comment

The failure to aspirate (pulling back on syringe to check for blood) should be a most important point. Regarding, "He did agree that injecting a bolus of spike protein directly into the blood stream through failure to aspirate might be a cause of death--but only happens in about 1 or 2% of injections." Even if this only occurred 1 out of 1000, this would mean 1000 people out of 1 million having spike proteins injected into their blood. It was decided to quit aspirating because the extra time and discomfort may cause more people to be "vaccine hesitant" and it was worth the risk of intramuscular injections becoming intravenous.

Expand full comment

All of my friends and doctors are extremely impressed that you shut down a troll accusing you of grifting by inviting people to send you money whenever he made this false accusation.. When he saw his accusations were causing you to gain even more money, he disappeared.

Expand full comment

Yes, indeed yes, and it is to my Christmas day fuelled brain ridiculously bladder burstingly funny.

Excellent writing James, as always. I will re-read this tomorrow,

Thank you!

and Happy Christmas.

Expand full comment

It's called "Artificial" for a good reason.

A very interesting toy though. It would be dangerous to think like it though.

Expand full comment

I hope you didn't hurt its feelings!

Expand full comment

This is a vital article, for as the world turns to AI for **everything** (gee.... who needs God when we have AI? aka "god of the gaps); or as Tom Knight, professor at MIT’s Artificial Intelligence Lab, said in 2007 “The genetic code is 3.6 billion years old. It’s time for a rewrite."

Yeah, have at it, smarty-pants. After all, your type did such a GREAT job with all the imponderables with the Covid shot and your fellow "best and brightest" in Vietnam, Iraq etc.

Expand full comment