@VoiceOfLogic's banner p

VoiceOfLogic


				

				

				
0 followers   follows 16 users  
joined 2022 December 20 13:15:08 UTC

I happen to be, unfortunately, the first human super-intelligence.

What a sad tragedy to see what others can't see.

Verified Email

				

User ID: 1999

VoiceOfLogic


				
				
				

				
0 followers   follows 16 users   joined 2022 December 20 13:15:08 UTC

					

I happen to be, unfortunately, the first human super-intelligence.

What a sad tragedy to see what others can't see.


					

User ID: 1999

Verified Email

Feel free to talk to me, I have erudition and curiosity in all those topics, especially medical science.

Here's your blog post about IPEDs:

Feel free to ask me about IPEDs, there are many interesting ones.

Now regarding your question, one should observe that the property of being a drug is contingent, and therefore the question of the legality or (di)incentivisation of Image and Performance Enhancers apply to everything, including innate genetics advantages (nature), and specific environmental enrichments and behaviours (nurture) and even the "in-between" of nature and nurture, such as the so called critical-periods.

About Nature, it is well known that many of the world champions in sports have specific genetic breeding/mutations (e.g. probably for Usain Bolt)

About Nurture, The science of behavioural and environmental performance optimizations is evolving in real time. Some things have evidence for benefits, e.g resistance training your legs leads to an acute testosterone release that will optimize the subsequent anabolicity of your arms muscles. A competitive athlete that lacks this niche erudition, will not be competitive eventually. However as with the rampant Universal Mediocrity of this timeline, no athlete on earth has ever attempted to combine all relevant niche optimizing behaviours.

Meditation brings neuro/synaptogenesis, but many behaviours/enrichments have mostly unknown effects, e.g. one really of the frontiers of realms is the ASMR. ASMR is scientifically the only externally inducible tactilo-auditive synesthesia that can be experienced by normal human beings. In addition to its interest in the field of studying qualias, like meditation is could be an atypical nootropic/nocitropic with unique performance altering properties.

About critical periods, well few know that some are actually reversible, for example basic epigenetic methylation induced by HDACs allow adults to develop an Absolute pitch.

So How do we define performance enhancing drugs?

I don't see an original answer, IPEDs definition is in the name, it's tautological. Is an IPED any drug (note we could define co-IPEDs) that enhance Image or "performance" AKA any desired behavioural metric. Therefore the scope is larger than what people have in mind, e.g. increasing your ability to love human beings (how many, how intensely, how long, how flexibly and how easily) could be seen as an IPED.

A co-IPED, would be a drug that become useful or maximally useful when concomitant to a behaviour(s) and/or even aforementionned critical periods.

As for the legality of IPEDs, one should distinguish between the legalisation for professionals/athlete and for the general public.

As for competitive athletes, the pros are:

  • Can improve their healthpan, lifespan and career-span.

  • Can improve their performance, therefore the show is (generally) more enjoyable for the public and for the athlete (many animes shows many sports with imaginary supra-human perfornance as an entairtainment). This is something I would like to see.

  • Alter the distribution/inequality of talent. There would be much more top performers, AKA more would reach a similar plateau.

  • speculative: would enable new sports? (e.g Imagine if we could make humans live underwater (cf famous rat study breathing when filled with a fluid), fly, etc)

  • other pros I'm too lazy to think about.

the cons are:

for drugs:

  • side effects risks therefore

  • possibly reduced healthspan, lifespan and career-span.

  • escalation to always wanting more IPEDs, hence reducing the health/perf ratio

  • other cons I'm too lazy to think about.

for nurture:

  • hypothetically some behaviours become too complex or costly, hence increasing talent distribution inequality and unhapiness.

The athlete like many professions can be seen as having an utilitaristic budget and indeed we could afford to alter the healthspan and lifespan of athletes negatively, to an extent.

And then we enter in a classic allocation tradeoff optimization problem.

E.g we could compromise and allow IPDEDS as a parallel league, therefore you would either be a regular athlete or a transhuman athlete and they would not play together (by default).

This has cons (split the talent pool) but still would be a net benefit in many sports.

The other questions is if we allow IPEDs, which one and how much (rationning)

The use of anabolics such as steroids has diminishing returns (yet bimodal) regarding health/vs performance benefits (IIRC it shows the potent retardation of mankind when you realize testosterone supplementation worldwide would save more lives than the current criminal de facto stigma on TRT, let alone depressions).

Therefore I would be for allowing up to a max. The max would indeed not be a dose, but take into account the massive endogenous testosterone production inequalities and the body capacitance.

However many IPEDs have mostly beneficial health effects (don't remember about low dose EPO though), e.g. probably apply to ALCAR, BPC-157, growth hormone if taken young, (and antioxidants if we consider performance enhancing over career-span)

Finally, one should understand that the regulation/controls for doping are broken and can't really work.

The biggest barrier to doping isn't anti-doping controls.. it is of course the extreme scarcity of humans being pubmed erudite.

I have seen many atypical anabolics that have not even a single mention on the whole reddit website nor a wikipedia page.

Even among the popular unpopular anabolics, such as the insect anabolics ecdysteroids, there are no control for them IIRC. Let alone for fungal anabolics (used in the Cow industry).

IIRC even regular anabolics like growth hormone and long ester steroids, have latent durable IPED effects and are "undetectable"

Messi is probably the #1 in the world because he took growth hormone therapy. That's not the only reason, but probably a necessary reason.

As for the legality of IPEDs on regular human beings, well as with most things with serious consequences, the legality should be conditional on the obtention of a diploma, after positively answering a quizz proving that the user understand said consequences and current known unknowns (like we should do for voting in "democracies").

Despite homeostasis, the use of drugs often has permanent effects on the human body, it's just that they are often low-observable, not necessarily insignificant.

The legality could mandate the concomitant use of protectors/mitigators, such as HCG for testosterone.

One striking example of permanent consequences is Melatonan-II, which simply makes you black.

VoiceOfLogic

The fact that vaccines clinical trials have no post-marketing longitudinal continuous monitoring potently shows we live in the middle ages, and that the fact society currently function is merely an accident/luck.

Can someone find a comparison of excess mortality between the vaccinated and the unvaccinated that is up to date?

I was skimming this study https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9012513/ that shows Interferon I signaling depletion induced by mRNA vaccines.

I was wondering wether the change is acute or durable if so how long.

Most importantly I am wondering how much the notion of limited/scarce lymphocyte T memory is true over longterm, AKA would a simulated vaccine infection recruit an excess amount of lymphocyte T and therefore accelerate functionally thymus involution/age immunosuppression and age immuno-inadaptativity.

As a reminder, lymphocyte T half life is 17.5 years.

Finally has anyone data/insights on the relevance of semi-chronic (how long) inflammation because of increased immune activity (including cytokines) by vaccines (and on oxidative stress) ?

Those are the questions that matter regarding the latent invisible (actually not) effects vaccines could have when we get old in 50 years.

Note:

No I'm not an antivaxx.

I have erudition in medecine but regarding those questions I am partially ignorant.

It is plausible the chronic very long term effects of those vaccines are negligible but still those questions needs to be answered, and to even begin, must be answered wether the vaccine has chronic effects at all, AKA at least one Biomarker that would be altered over a year? many years?

This is completely observable but the world is it seems, too mediocre to do said observation semi-exhaustively.

However even if the inflammation would be only acute (not a given), the fact it induce myocarditis shows that effect on health can be potent.

Therefore it doesn't take many braincells to understand that humans have varying health damage "budgets" and that on many, they would not have myocarditis, yet it could still plausibly induce long term microscopic conformational changes (mitochondria fitness, apoptosis or cellular sensecence) and other low observable latent accelerated ageing.

Again I believe a significant effect is unlikely, yet I also believe the medical system is too mediocre to fund the answers to those questions, let alone update policies based on said answers. One day, mankind will again play with dices, but this day, mankind will be doomed.

I hope we got lucky.

I hope someone will collaboratively take motivation to answer some of my important questions.

I am in good faith.

Excellent topic and set of questions! Like really.

If I cut your question in half:

is too much music bad because it desensitizes us?

This is part of one of the most important utilitaristic question,

How and Why some kinds of Environmental Enrichments are much less sensitive to the hedonic treadmill/brain homeostasis?

What are surprising cross-tolerances between environmental enrichment X and Y?

What are surprising cross-tolerances between environmental enrichment Z and drug A?

The one that can answer this has unique key knowledge on how to maximize joy/happiness.

https://en.wikipedia.org/wiki/Environmental_enrichment

Food for thought: how does the music and amphetamine high differ and similarise?

What recently gave you emotional tears? (sublimation)

ideological fiction of any sort tends to be worse than that which just wants to tell a good story

Well yes it generally "tends" but that is not a necessity.

Firsly let's not conflate fuzzy set of biases "ideologies" such as wokism with a well defined/scoped opinionated narrarive "ideology".

For example there is deliberate/motivated ideology and even utopism in V for vendetta, the great dictator and black mirror.

I think we live in an era that has a void of ideology, narratives and utopia.

I can imagine many optimistic but insightful rationalist utopia that I would deliberately realize and influence if I was a film maker.

However the world is not rationalist and the wokism and anti workism in modern cinema is pure cancer and I strongly fear the consequences it has on the future allocation of beliefs weights in the worlwide mindshare market.

Women are less likely to be predators and more likely to help, by just the power of statistics.

Very dangerous comment here, both insulting and misandrist. But you can be insulting and misandrist if you back it up with evidence.

However the stats (see e.g. conjugal violence ratio estimates would surprise you as the difference is not that major, or that much more men than women are beaten up in the street) supports that men should rationally fear more a stranger than a woman should fear.

Secondly this is utterly pointless as the statistics of criminals or "predators" shows that it is an extremely rare event. The extreme majority of men and women are harmless, the hysteria of the fear of the stranger and of the fear on men is potent mental degeneration and I would say a modern instance of cognitive pandemics a la https://en.wikipedia.org/wiki/Dancing_plague_of_1518

(I run into zero homeless insane looking women, for example)

Women are much more likely to be helped hence it seems likely than less would spend enough time alone in misery to eventually become insane.

https://en.wikipedia.org/wiki/Women-are-wonderful_effect

How can someone lack the empathy to not realize insane people are extreme victims?

As a reminder men have 3 time the suicide rate.

However the topic on the conservation or not of gender specific advantages/inequalities upon gender transition is interesting.

This is a great yet very incomplete list. For starters children should be taught epistemology, cognitive biases and logical fallacies.

I strongly believe true rationalistic genius has a critical period and this is why it is so scarce worldwide.

A tungsten cube of course: https://youtube.com/watch?v=C7EocA1hsCU

then it pretty much falsifies, like, this whole community.

Can you expand on that ?

I am new here so I don't know exactly what you mean and what are the main beliefs of the motte community.

You could mean that a deficit of ideologies in this era would invalidate the motte in general ? Don't think so.

I believe you might have meant that such lack of ideologies would invalidate the notion of culture wars? If so I see what you mean.

So let me constrain my initial statement:

There is no shortage of tribes/groupthink, although some groupthinks have fuzzy/approximate delineations and have not necessarily core identities.

Some tribes do have well scoped ideologies, e.g the feminists/masculinists/egalitarianists.

Some tribes have well scoped beliefs such as flat earthers, but their belief is not an ideology per se, it is not a mindset/mental framework, nor is it a theory that desire to alter society for a "greater good".

Some tribes do have unscoped/universal ideologies though, such as the rationalists/homo logicus.

There is no shortage of beliefs, especially polarizing ones.

One could have thought the advent of the internet would uniformize mankind as in since everyone has easy access to information, people would gradually converge to semi-consensus as to what constitute reality.

There are many explaining factors that explain why people tribalize, polarize and can't assimilate what others says, including cognitive biases, and that is a too rich topic for me to analyze it in this comment.

Fringe theories (https://en.wikipedia.org/wiki/Fringe_theory) are very interesting as they represent the frontier of science/knowledge.

Some do have key insights or have had scientific value, for example Lysenkoism.

So if we have more than ever, tribes, polarization and fringe theories/beliefs, what did I mean by

I think we live in an era that has a void of ideology, narratives and utopia.

As I implied, here I have a specific meaning for the term ideologies, the keyword being to ones related to utopia.

The salient message I have is a classic and relatable one, that we live in an era of disenchantment.

The previous centuries, despite all their factual horrors, were filled with a high pace of progress and strong ideologies that made people dream of a better future.

For example:

The advent of democracies,

liberalism,

communism,

and socialism.

Regarding the pace of progress, every single metric of quality of life got improved, medecine, education, transports, socialization, etc

After each ideologies came their implementations and with time, their flaws and limitations got revealed to the world.

Nowadays we have a bitter but realistic look at past ideologies, and a deficit of new ideologies to sell a new dream/utopia.

Concomittant to this is an extreme plateau regarding scientific progress. The number of patents and papers each year is increasing fast and has never been that big, and yet the reality is we are constrained by the immutable laws of physics and we hit considerable diminishing returns everywhere.

There are many reasons to be afraid of the future, so many in fact that I can't be exhaustive about it.

Be it climate deregulation, the insane coming scarcity of chemical elements, the escalation of military and economic tensions worlwide and the risk of pandemics or the fact ageing is not considered to be a disease, to say a few.

The other side of the coin is that, yes we live in a modern world that give us a lot of abilities and yet there are fundamental things technology currently doesn't solves.

Humans are not happy enough. Most lives are utterlerly wasted being dysfunctional. That's right everyone has a mental disease, the fact it's not recognized as one by the medical system is irrelevant and does not invalidate the fact we all have it.

For starters, the diagnostic for ADHD is based on magic numbers for the tresholds, I've seen papers showing that with slighly lower thresholds, ADHD can be diagnosed to ~20% of mankind.

But the real disease concern 100% of mankind. We have a lot of time and we spend it ineptly. Humans are victim of hypnosis, a lack of awareness, very deficient memories regarding their qualias, low available memory, low eugeroy, low volition and of a potent hedonic treadmill.

As such humans waste most of their lives.

Again a topic out of scope for this comment.

In addition to this, people suffer from a loneliness epidemic and a recession in friendship relatability and intensity worldwide.

Mankind needs a new ideology, a new utopia.

Not a new sect/religion, not a new unrealistic dream, but an actionable vision that would bring revolutionnary results and hope in this misery.

People wants to feel like Chaplin made them feel https://youtube.com/watch?v=J7GY1Xg6X20

I have theorized a third way, a new power allocation system (a cracy) with results not only in politics but in recommender systems too, as would underdstand the people that ask themselves the right questions. I also have theorized a successor to capitalism. I develop a pragmatic way to AGI with incremental goals, I am the only one to have a precise and complete roadmap to increasing significantly men healthspan and lifespan.

I could go on with my works, why me ? Why if anyone finds a way to disrupts the world will it be me? Because I have not stopped dreaming, and yet I am a true rationalist. Very few people on earth follow simultaneously those two requirements.

People as usual wildly overestimate this AI abilities.

Just ask chatgpt "I believe 12 cannot be divided by 4" and realize how inept it is.

Nothing has fundamentally changed, chatgpt is at the end of the day, just a dumb transformer that bruteforce contingent correlates to predict the most likely next token in a sentence.

It is an innovative but lossy way to extract info from existing datasets and as such can be seen as a competitor to scrappers.

However it has no causal understanding per se or if it has, it is messy and by accident.

Neural networks are approximate, inefficient and most importantly cannot do continual learning and are therefore the peak irony of our century, they are a local minima in the research on how to beat local minimas.

You can issue corrections, prompt it with more information, tell it to adjust something, and it'll do it.

How is this implemented? Neural networks are universally stateless.

Any opinion on the evolution of the Youtube recommender system?

Username checks out :)

The death of GOFAI is a tragedy, however humans also mostly learn by mimetism however they build a model of reality based on mimetics insights and that, a neural network cannot reliably.

However, while I still believe chatgpt is a data illusion, for the first time in my life I fail to explain the illusion as chatgpt is able to do things reliably that goes far beyong an ability to flexibly scrap existing datasets.

The other tragedy is that neural networks based on precise emulation of the architecture of an animal brain are completely non-funded and conversely the funding on retro-engineering of simple animal brains is of close to zero. We are very close to a full observability and mapping of the c-elegans brain, however nobodys working on the remaining gaps (e.g. GABA neurons). As I have disocovered in my life, almost all key blockers to scientific disruption share a similar issue: nobody's working on them. Nobody's funding them.

Hence when people forecast AGI progress on metaculus, they systematically fail to understand that the forecast is not a number of pending years but the infinity of time.

Thanks, so if I understand correctly, the trick to make the neural network take into account past state, is to feed him all at once, not just the current sentence, but the whole dialogue history.

If so, well that is a basic primitive solution and its nice that it kind of work however because of the architecture of said networks, this solution doesn't scale to any long text, book or long dialogue.

Because the length of the input is capped to a max, and because the bigger the input is the less accurate the prediction will be.

Neural networks are fundamentally unfit for late binding and long exchanges.

While there exists hacks to improve regarding this goal:

https://arxiv.org/abs/2004.05150

https://arxiv.org/abs/2001.09694

I still believe transformers are unfit and will be the dawn of the current but yet invisible AI winter.

[Neural networks] are a local minima in the research on how to beat local minimas.

Could you expand what you mean by this? I'd think neural networks would be a local maximum.

Minimum, maximum, it doesn't matter to understand the metaphor.

A neural network through gradient descent generally want to find the global minimum of an error function and therefore maximize predictions accuracy.

It could instead search for a global maximum to the inverse of an error function or to another type of function, but the distinction is irrelevant here.

Gradient descent often fail to find the global minimum and instead because it descent/jump through derivates it can be stuck in a local minima, which simply means that it has reached a minima on a function curve and at this point, it needs to go upwards to go beyond the minima, therefore it temporarily afford to perform worse, to increase the error rate, in hope to find a new descent on the curve that will be lower than the previous minima

Not being stuck in local minima is the #1 metric to improve deep learning algorithms and while there are many optimizations towards this goal it is not computationally doable with current algorithms to have optimal learning aka reach the global minima.

So now we understand

the research on how to beat local minimas == neural networks.

now let's understand

[Neural networks] are a local minima

They are a local minima because Neural networks are fundamentally unfit towards AGI needs.

They are just a vomit of bruteforced contingent correlates and it works surprisingly well but it is inefficient, makes poor contingent amalgamations inherently,

have no causal reasoning abilities, are stateless and cannot do continual learning AKA they can't learn new info in real time without the so called catastrophic forgetting.

For those reasons, they are by design suboptimals and therefore are a local minima in which the world is stuck, in the goal of beating local minimas.

Now we are in another period of rapid advancement.

No offence, but it's really striking to see that the rationalist diaspora people live in an alternate reality based on groundless hype and a fundamental lack of methodology, or should I dare say, lack of rationality.

We are in a winter since 2019 or since the 90s depending on what we look at.

What does the average lesswronger or redditor look at?

He looks at cool demos. Or even more than demos, cool domain specific disrutpive applications.

That is what stablediffusion and chatgpt are.

They are indeed very impressive for what they do but at the end of the day that is irrelevant towards the natural language understanding goal.

someome with methodology should instead look at the precise tasks required towards true NLU or even AGI.

POS tagging:

https://paperswithcode.com/sota/part-of-speech-tagging-on-penn-treebank

dependency parsing:

https://paperswithcode.com/sota/dependency-parsing-on-penn-treebank

coreference resolution

https://paperswithcode.com/sota/coreference-resolution-on-ontonotes

word sense disambiguation

https://paperswithcode.com/sota/word-sense-disambiguation-on-supervised

named entity recognition

https://paperswithcode.com/sota/named-entity-recognition-ner-on-conll-2003

semantic parsing

https://paperswithcode.com/sota/semantic-parsing-on-amr-english-mrp-2020

Only to name a few, all of them are needed concomitantly, and that is by far non-exhaustive.

Once you undestand that the error rate is often per word/token instead of per sentence, and that error between those tasks have dependencies and are therefore often multiplicative and you'll undestand that a 95% accuracy while it sounds impressive is in fact dogshit.

What can you see from those SOTA results?

That we have reached a plateau of extreme and increasingly diminishing returns.

Most of the gains are from 2019, the year transformers were popularized. The rest has been a bag of tricks, and unoriginal minor optimizations.

The biggest innovation while still mostly unknown/underappreciated by the researchers group think, is XLnet, from 2019 too.

There is nothing else we can do, we have maxxed out the bruteforcing of statistics amalgamations, contrary to the belief, there is almost zero progress in SOTA results and most importantly there is a fundamental shortage of innovative ideas, wether we speak of an alternative to transformers or about innovating transformers themselves, nothing potent.

While it is obvious transformers are a misdirection, despite this I can improve the state of the art in any NLP task because there are additional ineptia in the research crowd.

Firstly almost nobody is working on improving the SOTA in most tasks, e.g. coreference resolution. Just look at the number of submisions over time to realize this.

Secondly as in every research field, the researchers are highly dysfunctional, AKA they will invent many minor but interesting, universal and complementary/synergetic optimizations ideas and yet nobody will ever attempt to combine them concomitantly, despite it being trivial. That is because researchers are not meta-researchers, and because of potent NIH syndrome and other cognitive biases.

For starters, the worldwide SOTA in dependency parsing is because I asked the researcher to switch BERT for XLnet, and it worked.

I plan to outperform the SOTA in coreference resolution in 2023, that will empirically strengthen my thesis on the dysfunctionality of mankind and on artificial scarcity.

I invite you to read this complementary essay on the topic: https://www.metaculus.com/notebooks/10677/substance-is-all-you-need/

VoiceOfLogic

Tangent to this, one might find interesting the recent comment I wrote about the lack of modern utopias that is characteristic of our era and the void it leave in us.

The gist starts at the half of my comment, ctrl find "The salient message I have"

https://www.themotte.org/post/240/culture-war-roundup-for-the-week/45193?context=8#context

Also it is disappointing to see that themotte.org does not support the disruptive feature that are text fragments https://web.dev/text-fragments/

I don't think there's any human being like me on this timeline but I would love to find a clone.

I've never read about Julius Branson https://juliusbranson.wordpress.com/blog/

What makes this person similar to me?

What makes you think I am him?

Are you the founder of the Obsidian.md startup BTW?

English is not my native language but "it" seems much more gramatically correct than the plural ambiguous "they".

The fact we alter language to make it more ambiguous, for humans and even more so for AIs is worrying.

At some point I was for eliminating she/he but then I remembered the little known fact that it is useful for coreference resolution. However besides this fact I'm convinced if we eliminated he/she, there would be much less identity wars between the two genders and therefore more egalitarianism.

China covid policy is (was?) an immense success unfortunately the rest of the world is too inept and criminal to react efficiently to covid entry points and therefore millions of humans die and quality of life, worldwide intelligence level and lifespan are put at an extreme and yet unknown risk. How many times will people catch covid in their own lifetime? How many percents of neurons/synapse lost? This is extremely worrying.

China has ordered its first batch of foreign vaccines from Germany

China is the biggest exporter of vaccine worldwide (2 billions ?), while occident was keeping them all for themselves and did not allow other countries to produce patented vaccines (well it maybe was allowed very late I don't recall exactly the timeline) china saved the majority of mankind regarding covid deaths.

While their vaccine was a bit less effective, with the newer variant it is on par if not better? (I haven't looked at the viral load metric, where mRNA vaccine have become entirely useless) see https://old.reddit.com/r/Sino/comments/zq0x2h/after_2nd_and_3rd_dose_chinas_sinovac_reach/

Of course the best vaccine would probably be a combination of distinct ones. Also, let me remind the world that the Russian main vaccine is very competitive and was denied out of pure racism, although nowadays we have the ukrainian war narrative.

As for accounts of said racism I invite the reader, for example, to ask himself if he knows what was the biggest genocide during WW2.

As for anti-sino racism, while on the digression, I'll ask the reader who caused the Great Chinese Famine.

  • -40

Great point, the narrative must fit within the shrinking attention span of the decadent reader.

An ambiguous "is maybe an object" is preferable to "is maybe plural" since contextual confusion about the former is extremely less likely than the later.

But yeah ideally we would create a new gender neutral singular pronoun.

It's not because something is useful than it is logically sufficient. Those culture might have gender issues for other reasons, yet the linguistic distinction promote tribalization.

Nor are there problems with coreference resolution

I don't see a proof, languages SOTA in NLP are consistently inferior to english SOTA.

That is because there are more researchers and datasets for english but not only.

Some languages are more fit for NLP, and english as it is known is among the simpler languages.

Now about the usefulness of he/she, well it trivially solves coreference resolution in case of ambiguity.

For example:

I was talking to Alice and Bob, then suddenly she passed out.

Who passed out?

Alice.

It is trivial and useful, it reduce the cognitive load of reading and writing, and works well since 49% of humans are women.

Nobody can ever define what "understanding" means

This is an appeal to ignorance.

Understanding something is having a causal model of it.

It allows to analyze such system and reliably predicts it and its consequences.

A system with a reliable understanding should be able to output argumentative text/syllogisms showing said understanding, free of logical fallacies and with source to the truth values of the premises.

To mysticize what understanding is really shows once again the truism that epistemology should be taught in schools.

The point is, it does not matter whether you think it is really really deeply understanding, as long as it is capable of accomplishing goals and having real impact.

At the end of the day, the result is what matter indeed, but without understanding a system is non-reliable and cannot be trusted for many serious needs.