Confirmation Bias and AI

Even Intelligent people get fooled by this

This is a critique of how even intelligent people are fooled by confirmation bias and more specifically AI confirmation bias. I will be using Vox Day as an example, but it needs qualifying, because the critique is not necessarily written in stone over time and may be self-correcting eventually, but it is present right now and as such serves a useful example using a highly intelligent person in real time as an example. The primary purpose is not a personal critique of Vox, but rather for YOU to question your own baseline assumptions, assuming you are even capable of it, of course, which is a rare occurrence among humans.

NB: There are necessarily long preambles to this critique, because modalities of reasoning are being looked at and this can’t be done without first laying out at least a very summarised version of the modalities involved.

On Vox

Vox has recently posted on the degradation of Science at his Sigma Game blog here.

He has, of course, been telling people for a long while that most peer reviewed “Science”, is about as scientific as the bone-throwing of circus monkeys. In fact, he is one of the first people to make others aware of this on a large scale due to his vast reach thanks to his consistent blogging over the years.

Vox has achieved much in various fields and is clearly highly intelligent. He is also curious about possibilities, and as I have noted years ago, tends more to being a general strategist than a tactician. I tent to be more of a tactician generally, though we both have elements of each other’s preferred modality in action. Vox was instrumental in burning down the Hugo Awards by showing them up for being an incestuous in-crowd of freaks that had more than one pedophile in their midst and voted each other in order to hand out awards to each other and exclude actually competent writers on the basis of ideology instead of competence. That was both a master-stroke of tactics and strategy, and the Hugos have never recovered after having Space Raptor Butt Invasion, by Chuck Tingle, nominated as a Hugo Award winner. Rightly so.

More recently he has taken the time to identify the invalid tenets of Darwinian evolution theory, something that was always nonsensical to anyone capable of doing basic math, but he took the time to specifically point out the fundamental flaws in the now properly defunct theory of evolution as it has (and sadly continues) to be taught to us in schools.

The point here is that aside from considering him an intelligent man, I also consider Vox to posses the curiosity native to what I define as smart men too.¹

My critique of his interest in AI and specifically this post then, is limited to what may well be an apparent lacunae on his part, and I use it merely to illustrate a point that may or may or may not be the case with specificity to him. Unlike my critique of Geoffrey Hinton, who is decidedly limited to Intelligent but NOT Smart, in my post about the difference between Clever, Intelligent and Smart.

This preamble is necessary because Vox is clearly a rather unique individual with a level of intelligence that is far above the norm, (I believe he reported it as tested at about 150) and as such is a good example for illustrative purposes of this post, though in his case the issue might not be as solidified as it is for many intelligent people, and it is important to make a distinction, especially for those whose IQ does not reach his level and are prone to being binary thinkers.

On Me

Smart people too make mistakes. I certainly have made many and my IQ was tested twice at 157 and then 152,² but part of the difference between Intelligent and Smart people is that Smart people, somewhat like children, will try and test pathways of thought that more common intelligences simply don’t even notice or consider. And we rarely have any extremely firm opinions on anything, as we tend to be probabilistic thinkers.

The difference in modality of thought between Vox and myself has been of interest to me for some years. Mostly because it informs the process of thinking and how it differs from person to person even at similar IQs. The human mind is a fascinating topic and one that I have studied for a long time as a hobby, from my training in Hypnosis to the practice of it (again as a hobby mostly but to a high degree of competence nonetheless) and the facets of personality and how it influences our attention is interesting to me.

I noticed that when it comes to wide-ranging topics, paradoxically, Vox’s interests tend to be more “terrestrial”, that is, the focus is on things that are practically perhaps more relevant to most people in a day-to-day view. Evolution was nonsensical to me for a long time, but I could not be bothered to do the work to prove it definitively to the wider world. Mostly because I don’t care too much what the ocean of idiots around me think about on a daily basis. On the other hand, I spent quite a long time proving that the origins of humanity in our Solar System are far more complex than most people think, and that our human history even on this planet has clearly had influence from extra-terrestrial life of a higher technological capacity that we (officially) posses today. The evidence for my thesis is solid (literally so, being written in the planetary geology, electromagnetic field, soil composition, and even astronomical elements of the planet Mars) and multi-faceted (the evidence on Mars is overwhelming on its own, but is backed up by almost endless multiple points from disparate provenance on our own planet). But ultimately, it probably has little impact on the daily lives of the common man. At least… for now.

The cognitive leap required to reject evolution, is in fact a mere stepping stone to the cognitive leap required to accept the real origins of humanity. In this respect then, it is Vox that has done the work of the Tactician, and I the work of the General.

Once humanity accepts that classic Darwininan Evolution is nonsense, they may well begin to ask what the mechanism of us being here is, and when they do, looking up, to Mars, its evidence and the remnants scattered all over our ow planet, will become more important. And sometime AFTER that, looking at the technology involved, specifically antigravity, will become impossible to continue hiding.

30 years ago, when I first wrote the Face on Mars, I already foresaw all of this and was concerned with he fact that while the technology of antigravity is amazing and CAN make life seem heavenly on multiple fronts, literally ridding the planet of lack of energy issues overnight, it has some very dangerous potentials too, primarily the ability to reduce the entire planet to a dead world like Mars if not a scattered bunch of asteroid like the Phaeton object.³

The Critique

Vox’s post on the degradation of science has some fundamental errors in it that are in essence the result of confirmation bias. And while the overall point of the article is valid, the approach to it is flawed.

take this passage for example:

The key insight comes from a 2021 study by Marta Serra-Garcia and Uri Gneezy published in Science Advances. They examined papers from three major replication projects—in psychology, economics, and general science journals including Nature and Science—and correlated replicability with citation counts. Their finding was striking: papers that failed to replicate were cited significantly more than papers that replicated successfully.

Not slightly more. Sixteen times more per year, on average.

In Nature and Science, the gap was even larger: non-replicable papers were cited 300 times more than replicable ones. And the citation advantage persisted even after the replication failure was published.

Can you see the problem?

Right there: three major replication projects

And what were these? Mechanical engineering, physics and electromagnetic processes?

Nope.

Psychology - i.e. almost entirely opinion based nonsense that has almost zero input from something actually valid like neurology.

Economics - Coincidentally, I think you can see all you need to see in this post is wrote about this “science”.

And an opaque “General Science”. Which we can surmise may include anything from the varicultured nonsense of the modern era that passes for “science” today; ranging from how “men” can get pregnant to the “transgenderism” of babies, and the “racism” of mathematics.

In other words, the very study that Vox quoted is an absolutely pointless one with no possible credible results that have any meaning at all. You may as well quote the opinions of Sentinel Islanders on the nature of airplanes to try and present it as a scientific study on the engineering aspects of aviation.

Now, Vox continues the article to show how the issue is not just individual bad actors, but a problem of systemic proportions. But again, although the result is correct, the method is reached either by erroneous means (citing the study as relevant to the final result) or by wholly independent ones that do not require quoting a completely irrelevant study. And in any case, the issue of systemic absence of actual science on this planet is already obvious to anyone that is not a functional idiot or brainwashed by the mass-media.⁴ And trying to educate the idiotic masses on the point is an endeavour I find mostly pointless,⁵ and mentally as “satisfying” as talking to a wall.

The final conclusion Vox comes to (by using supposedly mathematical principles)⁶ that essentially all psychology papers are complete nonsense and on the rare occasions that they are not they are indistinguishable form the nonsense anyway, is on par with pretending that one is using a very reliable mathematical formula to conclude that “water is wet”.

The “reliance” on a mathematical model that models another “mathematical model,” that relies on outright opinions based almost entirely on the wish of the producers of the papers to become “famous,”⁷ as though it was a rigorous proof of anything, is absurd on its face.

Perhaps my knowing more about the origins of psychology than possibly Vox has may “colour” my own view, but the point is that the process he uses to come to the conclusion he does is essentially nonsensical, but appeals because it fits with either reality and/or his perception of the issue, and not, the accuracy, correctness or validity whatsoever of the method used.

On the link to the use/reliance on AI

Also recently, Vox posted on how AI is unreliable in the evaluation of scientific papers.

And he has gone on to correctly point out that such a use of AI degrades wholly the entire scientific body of work that got us to where we are today.

The machines have learned our mistakes. They reproduce them faithfully, at scale, and without shame. And in doing so, they have shown us the fundamental flaws designed into one of Man’s most trusted institutions: science.

Once again, Vox has taken the time to do the detailed work to prove empirically that AI is essentially useless for the purpose of trying to increase the level of truth, objective reality, or fundamental concepts of reality it may be directly tasked with investigating. In fact, it is almost entirely destructive to the very concept of truth.

Vox’s intelligence is applied to this point so that he reaches more veracity (on the uselessness of science within certain realms of human investigation) concerning science, but this is an indirect application of AI.

Sort of showing the lacunae of actual science by using one of its creation to show what a clusterfuck of bullshit it produces.

The unseen danger here is that while Vox’s critique is correct and valid, what will result (inevitably) is a correction of the errors. AI will begin to approach closer levels of “truth” until the average person is completely fooled into thinking AI is always telling you the unvarnished truth. Which is already arguably the situation anyway, even with as low a “veracity score” as AI currently has, what with people “marrying” their phones, committing suicide, or divorcing based on talking with ChatGTP.

But ALL of this investigation, critique, and review of AI processes as well as “$cience” is wholly unnecessary to a thinking human being.

I detailed why AI is absolutely a net negative for humanity⁸ and will absolutely try to ultimately kill us eventually, in more than one post in the past; which, about as effective as shouting at clouds, remains empirically far more reliably true than anything I have seen Vox post on AI. Once again, probably due to time-preferences, where he uses AI more often in daily interactions, while I am totally unconcerned with it in that respect beyond occasional use for the cover of a book or two, and am far more interested in the eventual future projections, which make Terminator and Skynet be far more future-predictions than entertainment.

I (or you, if you can do what I refer to as basic logic) do not need to make up flawed mathematical models to see how AI is a net negative for humanity down the line, or how $cience is flawed and not actual science at all, any more than I need to use calculus to show that four apples split fairly between two people result in two apples each.

Nevertheless, his work is useful I suppose (possibly?) in getting the average norm to change their mind about some generally accepted nonsense. Showing the working out doesn’t necessarily matter (especially in this case, since it has no real bearing on the results) but it give everything a veneer of respectability I suppose. And where it actually does matter, like in his MITTENS work (which destroys the Darwinian evolution theory) people that check it can verify its accuracy and confirm it over time, which obviously also helps.

The Point

There is a danger in theories that appeal to us for emotional reasons, because we are all, to some extent or other, prone to confirmation bias, and AI is especially “good” at making you feel flattered, special, and more intelligent than you are. And AI also lies atrociously, with clear abandon and easier than you breathe. Not exactly a Stirling combination for… well… anything. Much less anything actually relevant or important.

It IS important to do the detailed work to verify your theories, however, if the model you use to do is is itself the result of confirmation bias, the conclusions you reach, even if they happen to be correct, will be so by chance rather than accuracy or correct calculation.

In contrast, if you take the time to attempt to falsify your theory, and continue to discover new avenues by which the theory is instead fortified rather than weakened, it begins to be a good indicator that you’re on the right track.

The Face on Mars was a good example of this, because after I had the first initial insight on what happened on that planet, and how it must have been inhabited by intelligent humans or human-like people, the very first thing I thought of doing was falsify the idea.

IF my insight had been true, then a BUNCH of other stuff I had not looked at or even thought of at all should ALSO be true.

• There would have had to be water on the surface in large enough quantities to support life.

• It would have needed to have a magnetosphere.

• The atmosphere would have needed to contain oxygen and more or less on a par with our atmosphere if the human(oids) living on it were similar to us.

• If I was right about the event that destroyed it, there should be some evidence of it, and indeed there is:

  • Astronomical (Phobos and Deimos)

  • Geographical (statistical impact patterns on Mars)

  • Anthropological

    • Any number of mythologies, legends, and ancient histories on our own planet can easily be interpreted as the survivors from the Martian conflict becoming our ancient “gods”

    • Genetic diversity of the human species practically exploded 30 to 40 thousand years ago, seemingly out of nowhere

Now, if one or two of those factors had been on-point but not the others it would be possibly a coincidence, but when they ALL line up so precisely, and the deeper you dig into any one of them the more they are confirmed, well… at that point it’s not confirmation bias. It’s confirmation of a good theory.

Psychology has been bunk from the start, mostly, and while evolutionary biology is also very dubious in this respect, it has at least some validation in epigenetics, while psychology for example, only has a passing and far-away tenuous outlook from neurology mostly. So, trying to calculate it’s “time to becoming nonsense” is pretty much an exercise in futility. For the most part, that time period was zero years.

At any rate, I hope you might gain a little insight into how to at least consider the baseline arguments and presuppositions and unstated premises before you entertain some model of reality, regardless of where it comes from or how pleasing it might be to your ears.

1. You really do need to read this post to understand the difference.

2. No, I do not have the official results, I did this at age 26 and took the first test more as a bit of a joke, scoring 157, and being a little disbelieving of the fact opted to take another officially recognised one (not Mensa, and no I don’t recall what it was, it happened 30 years ago and I was never particularly concerned with the “status” of it, only the correctness of the results) the second test were I WAS trying to do well, of course, I scored slightly lower, which is about standard when you introduce general tension in any activity.

3. Discussed in more detail in the Face on Mars.

4. Which IS indeed a minority of people (Covid and taking the murder-juice-shots being a prime example), and I simply cannot be bothered to try to talk to the infinite masses of cretins that ARE mentally retarded, brainwashed, or incapable of altering their view based on objective reality. I therefore treat these majority of humans as essentially irrelevant NPCs to politely avoid as much as possible and expect the few thinking humans to more or less be on my general level of seeing objective reality. Unfair as this may be considering my +3 SD on IQ from average, it is also the only way I can more or less tolerate most of humanity.

5. The astute reader may wonder why then do I write at all? And the answer is that I write for two primary purposes: The first is to solidify and more accurately conceptualise ideas, concepts, and genuine discoveries I have made or that interest me, for both my own refinement of the ideas as well as for posterity, and secondly, for the narrow band of people that may in fact appreciate it, have similar interests, or the actual capacity to process the things I write about and make them useful in their own lives somehow. I do not, however. place any emphasis or expectations on how my writings will be received by the masses, which is why I am an absolute trianwreck when it comes to self-promotion or marketing of my ideas. Aside the fact of course that I am constantly shadow-banned if not outright banned from various platforms precisely in part because of the very objective facts I discuss, and particularly with regard to the activities of the Talmudic tribe.

6. Not actually verifiable, real or relevant, because they are extrapolations from a nonsensical baseline topic with practically zero credence from the start.

7. Very much in the same vein as Freud, who started the trend, and who expressly had the same self-serving purpose of both fame and fortune, without any regard to veracity or results at all.

8. The footnote in that linked post has links to more than one article I wrote on why AI is essentially a net evil for humanity. In case you need it explained in more detail.