The ONLY thing that AI Tech Bros actually created, and WHAT to stop about AI

because we'd really need some REAL AI, NOW.

This is not a rant against the types of software we misname as "Artificial Intelligence". AIs can and should do a LOT of good NOW, at least in all the ways I explained here, plus nuclear power, plus packaging, and surely much more.

This is not even a post against the few people I explicitly mention, who at the end of the day may be just the incidental champions, in a very "Professor Falken" way, of much larger trends and issues. They just happen to be perfect, if involuntary examples of some points I need to make. This is a post about the real nature of the currently dominant AIs pushed by the attitude which simply for brevity I call "Tech Bros", and about one thing that should be done about it.

This post originates from a video about AI by Patrick Boyle, in which he shared his surprise about "how these models can be successful at certain really complex tasks, but then fail at tasks you might imagine would be easy for them... They can do calculus, but then struggle to count.". That remark made something click in my mind, connecting in one coherent picture that observation, with many other, not directly related ones I had read before.

Side note, before getting to the main topic: please notice Boyle's observation that "AI will have an especially profound effect on professions traditionally requiring higher levels of education, whose work has been less affected by automation in the past." That is just the 9999....th confirmation of what I wrote about degrees: do get one, but only for yourself, NOT for any career.

One question, seven times

A righteous man may fall seven times and rise again, but the wicked shall fall by calamity, someone said a long time ago. So, let me ask seven times, about seven different falls, "How are the AIs produced by these Tech Bros?"

First time

Inability to see and do the simplest, most urgent things that matter in everybody else's everyday life is not intrinsically bad. It's what makes us human: had we lacked it as a species, we'd still be up in the trees. When it goes off the rails and sucks all the oxygen in the room, however, it's bad.

Tech Bros are exactly people who won't accept the world as it is, but not in the most mature way. Maybe it's because Tech Bros "get rich when they are still young and energetic... with plenty of time to try to prove [their ideas]" but after losing the humility to let those ideas evolve.

Tech Bros hope to "bring about the speculative fiction stories they grew up reading", pretending they were "science fiction, [when they are] usually fantasy". They can be found, as a book said about ten years ago, "straining, not always coherently, to grasp problems for which [they were] plainly unprepared."

Tech bros who are most emotionally unhealthy and immature members in the system become de facto leaders who shape the emotional landscape (that's not their fault, of course, just what happens in every industry since when Tech Bros were in kindergarten).

How are the AIs produced by these Tech Bros?

Their AIs show little evidence of any knowledge of "cognitive science literature", but many signs of a "wider problem of groundedness in real-world concerns" that makes them "fall behind humans on interaction with real world" and assert "wrong things as truth, with full confidence".

Second time

Tech bros are great at coding and electronics engineering, much less at everything else, especially social innovation.

How are the AIs produced by these Tech Bros?

Their AIs, Boyle notes,are "successful at certain really complex tasks, but then fail at tasks you might imagine would be easy for them... They can do calculus, but then struggle to count", or, as others said, "pass the bar exam but flunk preschool?". At a deeper level, their AIs appear to be lacking "a crucial aspect of human intelligence: rich internal models of the world", and therefore are really bad at things like "basic physics or social interactions-things that are second nature for people.

Third time

Tech Bros run an industry in which "the gender gap in AI is widening - not narrowing", also as a result of "not fitting into established male-dominated molds in AI".

How are the AIs produced by these Tech Bros?

Their AIs are mansplaining as a service that produces "stories that substantially reflected a brilliance bias towards men"" and seem unhelpful to solve the pre-existing, related problem that, even though men outnumber women in the workforce, those AIs expose 21% more women to AI automation.

Fourth time

Tech Bros are the folks who “move fast and break things", period.

How are the AIs produced by these Tech Bros?

Their AIs are angels who must be let lose, as soon as possible. Even if, borrowing Boyle's words, they are creating "a lot of confusion around [their usefulness, thanks to] a combination of people exaggerating their capabilities and [lack of] guidance on the best ways to use them [and the fact] that almost anyone can use these models with next to no training." Their AIs, that is, must be let lose, just like the addictive social media and smartphones by the same Tech Bros were let lose before, by regulators who haven't learned anything in almost twenty years.

Fifth time

Tech Bros religiously (yes, religiously) believe that Their Tech can always do everything for everyone, all by itself. They have always done that. From:

• the clueless Declaration of Independence of Cyberspace in 1996, the one that forgot that democracy can only be where our BODIES are...

• to Negroponte's One Laptop Per Child in 2005, based on "the arrogance and ignorance of the assumption that the presence of laptops would instantly transform young children into autonomous, disciplined seekers of knowledge and understanding... .. always evident [and] tragicomically epitomized when Negroponte famously said that he would favor flying over poor areas and throwing the tablets out of helicopters".

• passing through Zuckerbergs' vision of himself as "man of history... destined to be great.”...

• to finish (you wish!) with Marc "We had a problem of isolation, so we invented the Internet" Andreessen's unbelievably Solarian sermon last year…

it's always been one and the same glorious tradition.

How are the AIs produced by these Tech Bros?

Today, their AIs are, with the same religiosity presented exactly as an "unscoped system with the apparent goal of trying to do everything for everyone under any environment".

Sixth time

Tech Bros, like preteens and younger teenagers, are above contexts and nuances: “Having two identities for yourself is an example of a lack of integrity”, said Zuckerberg. Tech Bros build innovation that's too often fake, and sometimes (e.g. the Metaverse) is reactionary too.

How are the AIs produced by these Tech Bros?

Their AIs "assert wrong things as truth" with all the confidence of the most outrageous individual you can find in their social media today, and, just like those media are based on negating that humans need borders, risk to pollute and ruin themselves beyond repair. Their AIs are machines that can create "soulless facsimiles of "NOT innovative ideas", but cannot "make imaginative leaps and, at least for now, cannot even do it without consuming "internet-sized gobs of content" not to mention way too much water, energy and possibly even money.

Indeed, Boyle cites a study finding that, "while...subjects using AI produced higher quality ideas in the brainstorming sections of the study, there was a marked reduction in the variability of the ideas", which sounds a lot like the "less of everything" that Crichton said (30 years ago...) would end our species:

Seventh time

Tech Bros mostly belong to one political community (not judging here, seriously! Just observing a fact).

How are the AIs produced by these Tech Bros?

Their AIs are just like them, politically: "when I asked Gemini to decide whether Nazism or capitalism was worse, it equivocated and said it didn’t have any business making such judgments... Gemini refused to argue in favor of having four or more children, but it was happy to make an argument for having no children" (SOURCE). "Luckily" (please do note the quotes!) "uncensored and unbiased AI platforms that accelerate your mind" immediately came to the rescue:

The only AI Tech Bros have really created

I asked seven times, and the answer was always the same: what the Tech Bros have been able to create with AI, burning oceans of money and now demanding trillions more because they have nothing else left, is... themselves. Nothing more.

If "AI is a mirror" (Tim O'Reilly), it is first and foremost a mirror of its creators. The Tech Bros aiming to become immortal digital avatars haven't realized that too much of today's AI already is them.

The question, then, becomes, "why more of this?" Why, when there are actually good things AI should do, the sooner the better?

What to stop about AI and why

One year ago, some unhortodox Tech Bros asked to stop the training of more powerful AI system for at least six months. Personally, I think ten years would be more appropriate, but not necessarily for AI training. Probably, what should stop immediately, for years, is unrestricted, affordable access to AI applications (not research), but not for the reasons most people may imagine.

I don't say this out of the fake fear that "AI may kill us all ", as that may happen only if we were so terribly stupid to definitely deserve it. I don't even think that the block should happen to save students, or businesses. Making students dumb and slave to the toxic idea that the purpose of education is getting a great career is something that has been going on for decades, and the capacity of chatGPT to "write" essays just accelerates, not create. As businesses go, if there are managers dumb enough to deploy any tool just because it's fashionable, the sooner their companies die, the better.

I say "end AI access" partly because making certain tools easily accessible by everybody is exactly what made all the (bad) difference, and the more this continues at full speed, "the more likely it is to cause mayhem". The best but surely not only proof of this is the effects of AI on politics, with targeting voters with AI fakes and dead politicians campaigning from their graves being just the latest of many more examples:

Zombie politicians and fake voters, all coming to some election near you.

But above all, I say let's end AI access because very, very likely we just cannot afford everybody and their cat to run AI queries every second. Even the AI bubble will pop, eventually, by debt or because all the rank and file Tech Bros now making AI will figure out they're getting much less than they gave, and leave. Point is, can we afford to wait for that?

The first and most important reason to make AI inaccessible for the general public is that chips, energy and money are neither infinite nor cheap. The chips, energy and money that AI ubiquity would need should really go first to stuff that matters more, and is infinitely more important to do as soon as possible, both with and without AI. Stuff like this, or this), or this, for example. Please add your suggestions.

Usual final call: the more direct support I get, the more I can investigate and share content like this with everybody who could and should know it. If you can’t or don’t want to do it with a paid subscription, you may fund me directly via via PayPal (mfioretti@nexaima.net), LiberaPay, or in any of the other ways listed here.

Substack authors credits:

Thanks for inspiration to

Gary Marcus

and

Dana F. Blankenhorn

!