ChatGPTs

Most people nowadays dismiss philosophy as a joke, but yet we have something to say.

The ancient Eastern philosophy of the Mind (which goes back to Upanishads) identified the fundamental problem, which has given the name of “Maya” and defined as “veil” that obscures and hides the “real world” from the Mind (which is an ultimate Observer).

This veil, it is said, is made out of words (of bullshit) and wrong concepts of the Mind named (referenced) by such words.

Bullshitting of the rich and powerful (and of mere mortals) is the second oldest occupation in human existence.

Sectarian bullshitting already ruined what we call “science” (which is a particular evolved methodology for establishing what is out there and what isn’t).

Science, by the way, has been evolved as a reaction to endless torrents of religious bullshit, including theology and “abstract philosophy”, which provided a decent living for some and painful death to others.

In short, just accept everything what one hears is not intelligence but stupidity.

GPTs, however, by its algorithmic properties capture what is said most often, exactly according to the maxim “A billion flies cannot be wrong”.

There is a lot of crazy shit going on thanks to the social media and internet – doctors use “ideas” from the web or even based their diagnosis on what they have read at some crappy site that has the best CEO. Do you want to be treated by such a doctor? Noooo, for ourselves we want an evidence-based, precisely-measured individual medical treatment, thank you.

This is not a joke. All GPTs will give you pseudo-random, but absolutely convincing bullshit, and only real experts could even call the bullshit. It has already been shown that GPTs fail in all rigorous disciplines for precisely the reasons outlined above.

Just like mathematics is not mere statements of theorems and axioms, but the actual proofs being veryfied or actually being checked by lots of competent and highly motivated people for flaws and contradictions, and found none so far). This is our operational definition of what is True – what cannot be shown to be flawed, what stands (remains) after all the bullshit arguments has been discarded.

Notice that so-called modern science abandoned this golden standard and resorted to so-called “scientific consensus” which is a misleading meme for sectarian consensus.

Well, it is said to be OK to generate “templates”, “sketches” or tedious “boilerplate” like “unit-tests”, but again, even with unit-tests there is an actual danger of missing just one possible execution path or a check for an invariant, and the whole codebase would collapse one day.

The fundamental problem is that GPTs produce convincing bullshit, just like all the imposters do. This is a state-of-the-art methodology and technology for creating convincing bullshit and if being an imposer could be defined precisely, that would be how ChatGPTs select sentences for you.

Now lets get serious. What we have is another era of endless bullshit. Just think of all the bullshit all civilizations produced (and still preserving) about death and “life after death”. It was all very convincing bullshit, indeed, so convincing that Pyramids has been built in Egypt, and many other seemingly doing exactly the right thing “projects”.

By the way, one day, for the very same fundamental principle of sectarian bullshit, the Large Hadron Collider will become such a “Pyramid”, because all the “results” are based on the current sectarian consensus, and more importantly, all the instruments has been build on current assumptions and abstract mathematical frameworks. Just one single logical flaw would collapse and nullify most of the results, and I claim that such a flaw exist somewhere in enumerable pages of theoretical bullshit (they calculate and call it “measure” of what isn’t there).

The ancient veil of Maya (of flawed mental concepts) is very real. More real that many “results” of LHC has been produced.

Last but not least, programming, just like math, is hard because the whole conceptual build-up, without a single logical flaw, or a contradiction with What Is (or already known - already completed parts of a Jigsaw puzzle) cannot be skipped or short-circuited by any crappy GPT.

To run a bottom-up re-evaluation step (similar to what theorem-provers do) after a single change is required. A proper, sound scientific methodology would require to re-run an experiment (to confirm that nothing necessary was removed), which is costly, so no one does it.

Remember, that the Truth is what remains standing after all the bullshit has been removed. It is not the abstract Falsifiability, but True = not False. Literally. DNA is the actual genetic material, is probably the most striking example of the statement of truth arrived at by removing all the bullshit. GPTs, by definition, will never give you that.

GPTs just capture a snapshot of the whole current landscape of bullshit, with all its topology, so to speak. Notice, that just like any other snapshot, it is already old and not actual once it is made. The roaring sea of screaming bullshit is already different by the time you prompt your model. Only the most fundamental scientific results will remain unchanged, because they have captured something “real”.