Login
Please make THIS AI crash NOW
In case you didn’t notice, in the last one or two years:
95% of the organizations that tried generative AI failed to receive a measurable return on the investment. In 2025, 42% of businesses scrapped the majority of their AI initiative
AI chatbots put in place of humans to save some bucks have made:
air companies pay for lying to their customers
lawyers use made up court cases
therapists cheat their patients by using ChatGPT during sessions
AI-empowered scammers generated 75 Million nonsensical songs that took royalties away from human singers
AI-enabled teddy bears gave advice on BDSM sex and where to find knives
blindly outsorcing to machines activities like analysis and thinking is making most people lose, not gain skills that may be a bit more important and harder to re-learn than tasks that should be left to machines like e.g. digging holes in the ground
nominees for the 2025 AI Darwin Awards keep coming...
And as if all these “innovations” weren’t stupid enough in and by themselves, they’ll need so much electricity that even those who won’t get anything good from them will get higher power bills, and in some cases even lose their tap water.
Too much of a not even really good thing
If there is one thing that is painfully obvious to any real adult now is that AI like this for the masses is, in the best case, too much of a good thing, at too high a price.
Even if we could afford the financial and physical costs, which we can’t, the human species is not build to handle or produce so much data, so fast.
We already knew this from the Web 2.0 days, whose instantaneous, unlimited, many-to-many communication was supposed to free the masses both from ignorance (yeah, right) and tyranny (remember the Arab Spring?). Now, there are even studies arguing that “above a critical social connectivity, an explosive transition toward strong polarization must occur”.
Universal continuous access to fake human communications, services and relationships by means of artificial “intelligence” is just more of the same. Just too much for our collective health, human rights and productive political discourse, no matter how much good will is put in AI.
Sure, there are plenty of exceptions. That is, there are many gifted individuals who are doing good to themselves, if not to society, thanks to cheap access to AI. But they aren’t enough to justify mass unrestricted access to services that, one way or another, do dumb down or make humanly poorer the majority of their users. Unrestricted access to AI should be earned, not given away like candies or alcohol to Native Americans. Earned by merit and hard real work, of course, not money. Study without summaries some million words by humans who knew their stuff, write yourself some hundred thousand ones, and then yes, you may indeed start getting something good by a chatbot-generated summary. Hence my call.
This will end soon, but not soon enough
Luckily, this bubble is going to pop soon. If, as some experts think, we’ll be out of affordable electricity real soon now and increasing warnings like these are even just half right:
Oracle is already underwater on its ‘astonishing’ $300bn OpenAI deal
AI investment the only thing keeping the US out of recession
there are just very few years left before the billions of AI queries to fake dating profiles, homework, CVs, work that shouldn’t exist in the first place no matter who or what does it... all end.
But why wait? When the dust settles, we’ll all be better off, so why wait for something that will come, but may come late enough to not make enough of a difference anymore in our lives? Even if what today passes for AI wasn’t a scam, the sooner we stop unrestricted access to it, the better. Some investors will cry, but who cares? Collective mental health is much more important, and even the real economy, the one that should exist, would benefit.
Investors, I beg you: divest from AI NOW
Make the damned bubble pop now, please. The sooner the better (especially if you are a pension fund...). Besides mental health, the sooner this scam falls apart, the more money, energy and strategically critical rare earths and other materials are left to do stuff that actually matters.
The sooner this AI bubble crashes, the sooner saner public and private organizations can grab all the infrastructure it will leave on the ground, for pennies to the thousand dollars, and repurpose it for the kinds of AI we all really, urgently needs. I mean things like:
protein folding
help human law scholars and students to simplify existing laws without altering them
standardization of physical boxes
If you have any money invested in false, toxic AI, please divest it now!
Ditto for any “Consumer IoT” stock, of course
Last October, people were stuck in awkward positions in “smart” beds paid 2,000 dollars) because some computers thousands of miles from their bedroom went bad, or became unreachable. This broke the beds because they were designed to send up 16 GB of data per months to their makers, who could then share them with any advertiser or police force interested to know when and how people were actually sleeping in their homes.
Beds like those are just the last kind of totally dumb consumer IoT products that shouldn’t exist. Just say no to stupid beds and everything else that makes homes stupidly, uselessly “smart” now, and tell everybody to do the same. Both as a customer, and as an investor.
Usual final greeting: to support more writing here on this, or other important topics, thanks for considering a paid subscription or (even better) supporting me with one-time direct donations or in any of the other ways described here.
