The debates that cause debate over AI

AI has its fans and its detractors, and that's only natural. Technological revolution, the changing of an era, new uses, professional uncertainties, the end of the world, colossal fortunes, the end of diseases... We are promised everything and its opposite. Yet many questions seem to me to be poorly posed in relation to real issues that remain unanswered. And since a bad question always yields bad answers, I offer you a completely subjective detailed review from my perspective, along with proposals for new topics to gently bicker about AI.

AI IS GOING TO REPLACE US ALL!

As long as there is a human left with at least one hand on Earth, the ultimate solution will remain: unplugging the machines in the event of an uprising against humanity. And yes, AI is just like your washing machine, it needs energy to function—and much more than your washing machine, for that matter. Without energy, AI becomes as useless as a remote control without batteries. The race to build a multitude of mini nuclear power plants is just as crucial for AI players as the evolution of algorithms toward an ultimate model. As long as humans have things to teach algorithms, they will remain the master, the benchmark, the director. We will always need farmers, fishermen, livestock breeders, artisans, shopkeepers, athletes, teachers, artists, intellectuals, creatives, researchers, and I am forgetting many irreplaceable professions that, themselves, consume little energy…

The discussion topic I propose to you instead:

Are we all going to die in a nuclear explosion before we see ChatGPT 25? AI IS GOING TO KILL ADVERTISING, CINEMA, PHOTOGRAPHY, CREATIVITY…

We are promised that AI is going to kill all creative professions in the broadest sense. Ah, creativity, what a vast subject! Given that everyone has their own definition of creativity. Generally speaking, people agree that advertising is in a much less creative period than it was 20 years ago, as are cinema and TV, which face competition from series and influencers (less creative rather than more… wouldn't you agree!). And we are witnessing the same debates I’ve already seen with the arrival of every technology (before and after the digital revolution in advertising, before and after the Pixar revolution in cinema, before and after the resolution revolution in video games, before and after the arrival of synths in music, etc.), basically: it's not like it used to be, we used to do it like this back then, etc.

I just want to remind everyone that it is the human brain that is creative, not the tool, whatever it may be. And it is those who buy or judge creation who are conservative and censorious. So yes, we will be able to make AI creative or not through how we use it, like a pencil or a paintbrush. Creativity will remain a question of talent and the willingness to buy it. What far too many people call “risk-taking”—that of not doing like everyone else.

The discussion topic I propose to you instead:

Rather than harping on about the umpteenth death of creativity, let’s debate the multiplication of those who want to kill creativity, in whatever form it takes. It will be funnier, and we can pick fights with more people. Take this topic, for example:

The generalization of decision-making that satisfies no one: the disease of the century? AI IS GOING TO WRITE FOR ME, NO NEED TO LEARN TO WRITE ANYMORE

The problem with this stance: anyone who doesn't know how to write won't be able to write a prompt for AI to write for them. People will counter that voice commands, and therefore voice technology, will develop. That's true, but you still need a minimum vocabulary for the AI to achieve your goals—that is, to translate your thoughts. I am certain, on the contrary, that it is through culture that we will master the power of non-human intelligence.

I even think that whoever possesses sophisticated, virtual, global thinking—that of mathematicians, theologians, chess players, and philosophy professors—will be the dominant force in the next labor market cycle. If I understood correctly, one of the avenues for algorithm evolution that Yann LeCun is trying to develop is making AI understand a global physical vision of what it must learn and retain; that is, not a letter, a word, or an image, but an interconnected whole of all these elements—just like humans, in short. To understand the world is to dominate it. That has always been true, and it will remain so.

The discussion topic I propose to you instead:

Philosophy professor: soon a profession more sought-after and lucrative than investment banker? The professor would still wear the same old sweater, but his sweater would be cashmere. AI IS SMARTER THAN HUMANS

Engineers and researchers combine the potential knowledge of AI and compare it to a single human brain! Isn't there a little deception here? Let's combine the potential of 8 billion brains to compare it to the potential of a single LLM to make it a fair fight. We are told that the potential of AI is limitless, but that forgets that the brain's potential can be too. All it takes is to keep feeding it, developing it, and using it to its full capacity, like a muscle you want to build your whole life.

The discussion topic I propose to you instead:

What is the point of creating the world's largest virtual brain for dumbed-down humans? If man is as stupid as they say, he will use AI to invent new silly jokes, not to achieve universal happiness. TOMORROW, WE WILL ALL HAVE ROBOTS AT HOME

File this in the same category as "tomorrow we will all go shopping on Mars." I've already written this in another op-ed for LHC, but if investing hundreds of billions of dollars is meant to create a robot that will crack eggs for us and take out the trash, becoming obese with an IQ of 5 will become the global norm. All that just to sell a subscription higher than our cell phone bill!

We are predicted either the purchase of a domestic robot—inevitably reserved for the wealthiest, like all technological innovations (it takes many, many consumers before mass producing and lowering production costs)—or monthly rental plans for home robots. In reality, the military is already using robots in the field and AI for cyberdefense. Chinese police are already testing robots as deputy officers on the streets. To secure a home, just like a business, there will certainly be supply and demand. To assist healthcare workers in hospitals as well as at home, too. But I remain convinced that the development of robots will first take place largely through commerce and industry,