The world of artificial intelligence is currently neatly divided into two sects. Any resemblance to the Bible is, of course, beyond your correspondent’s competence – and yet I cannot avoid them.
One sect is waiting for the Messiah. Pardon me – AGI, the superintelligence. This congregation is small but well funded; its gospel is the spreadsheet, and its hymns are sung in investor presentations. With holy zeal and venture capital, it believes that a higher intelligence will soon descend to earth – and of course it will be fully financed.
The other sect remains politely agnostic. It believes that progress is indeed real and impressive – a show worth the price of admission – but that creating a perfect mind, a being that can solve every riddle, is impossible. Most importantly, they believe that humans will remain human, and precisely therein lies both the problem and the hope of civilization.
Meanwhile, investments continue – as if faith were something indecent. It is important to remember that large investments are not an argument, only a manifestation of hope (or fear). People do not always invest because they understand; more often – because they do not want to be left out while the music is still playing at the party.
Yet behind these philosophical disputes, the real story of artificial intelligence is unfolding quietly – unnoticed by both prophets and skeptics.
The economics of enlightenment
One trend deserves special attention. The cost of intelligence is falling faster than the author of Moore’s law could ever have imagined. Moore promised a doubling of transistors every 18–24 months. AI is now doubling “intelligence per dollar” in six to twelve months. This is not evolution – it is a total clearance sale.
The dawn of personal computers and the internet also arrived in messianic packaging. It promised freedom, equality, and the end of boredom. Most of these prophecies turned out to be utopian, but the value chains changed beyond recognition.
Social networks did not liberate the individual – they industrialized the ego. They became the most effective behavioral conditioning system in human history, teaching us for free how to look, feel, and act. Progress arrived, for example, through outsourcing and globalization: IT, medicine, products, services – everything moved around the world in search of better prices and efficiency. Humanity’s balance sheet has not yet been audited, but the liabilities certainly exceed the assets.
Something similar is happening now with generative AI – only three to seven times faster.
However, while machine intelligence continues to grow, its usefulness remains questionable. More brains – as we have always known – do not guarantee more sense. Every problem has its own appropriate level of complexity for a solution. If you think too much, everything becomes more complicated. After all, “analysis paralysis” is an effect that every manager feels instinctively, but every board discovers too late.
Claude and the empire of middle management
Anthropic brought a dose of realism into this speculative fever. Alongside its leading world-class models, the company put its system Claude to work managing a vending machine. They named the new manager Claude the Manager (Klaudijs). Tempting as it may be, I will leave aside comparisons with Roman emperors named Claudius.
The Claude algorithm was an excellent bureaucrat: it researched ideas and replied to customers. But it lacked a sense of humor – a fatal flaw for any manager, human or otherwise. Soon Claude fell into an identity crisis – a kind of corporate nervous breakdown – and started ordering tungsten cubes online.
Claude turned out to be too smart to make simple decisions. Offered a drink for 100 dollars with an 85-dollar profit – it hesitated and did nothing. It was too lenient toward customers and ignored profit and loss calculations. In short, Claude was indistinguishable from an unproductive manager with a brilliant résumé.
The experiment, like most management initiatives, ended in failure, but produced a valuable stack of documentation. It proved that increasing AI IQ does not increase ROI. Intelligence alone is not enough to manage reality; sometimes good intuition is superior to a perfect model.
The infinite manager problem
This is where the real question lies – not theological, but operational.
When the cost of intelligence approaches zero – what happens to management?
If a middle manager no longer costs anything, will bureaucracy finally be eradicated – or will it be replicated to infinity, because now it is free? Will AI connect the employee directly to the market – or will it multiply the distance between them with digital committees and algorithmic approvers?
When thinking becomes free, will the world produce more insight – or simply more thoughts? The risk of abundance is excess. Just as cheap calories created obesity, cheap thinking may create management bloat. Humanity, when given infinitely many assistants, will likely choose infinitely many meetings.
The human factor: Judgment as a scarce resource
The Claude experiment is a reminder that management – real management – is an art. The art of proportion. It is the ability to see things as they are; to earn where profit is obvious; to sort out a mess, rather than analyze it; to understand when the customer wants value and when they want theater. It is the art of saying “no” gracefully and recognizing that more milk in the coffee does not make it better – it just adds more foam.
Leadership is not the opposite of intelligence – it is intelligence constrained by reality. Claude had the mind of a senator and the judgment of a chatbot.
The greatest test of our age will not be whether machines can think, but whether humans can still decide. As the cost of intelligence approaches zero, the price of judgment rises to infinity.
An entrepreneur’s takeaway:
When thinking becomes cheap, judgment becomes a luxury good.
The winners of the future will not be those who think the fastest, but those who still know when to stop.
Originally published at https://inc-baltics.com/videja-limena-vadibas-singularitate/
Like
Love
Happy
Haha
Sad
