The Generative AI War: So now, what will happen next?
In March 2000, Cisco was the second most valuable company in the world behind Microsoft. The investors’ analysis then came down to a very simple notion: “If routers are essential to make the Internet work, then Cisco is the most important company on the Internet.” » A major analysis error, which could well be repeated with AI.
Many players are contenders for the leadership role in the AI of tomorrow. Thus processor manufacturers and computing centers, which justified that Nvidia (which is now both) is one of the seven companies in the world to be worth more than a trillion dollars. A perilous bet as new calculators with more breathtaking capabilities appear frequently.
EARN YOUR PLACE IN THE SUN
There are also the major AI models: Anthropic, OpenAI, Meta, Mistral… It’s a possibility. We can nevertheless observe that the level of competition in this field is unusually high, and we can legitimately wonder if everyone will manage to earn their place in the sun.
There are also specialized players: those who use these broad models to make them efficient in a particular sector: finance, journalism, programming, etc. There is no great risk in asserting that some of these players will experience spectacular growth, as the effectiveness of LLMs is powerful in certain areas, such as programming.
Here again, drawing historical parallels is interesting: at the beginning, large microprocessors did everything: cryptography, compression, imaging… but quickly, specialists appeared in each of these functions and the roles became segmented.
“ORGANIZE THE HOLIDAY”
As for the Gafam, they are certainly not going to be easily left behind. The big bet for several of them already consists of developing AI assistants: applications which will access the data of all of our applications, which will listen (with our consent) to our conversations and which will know us better than anyone.
We can thus entrust them with tasks such as “organizing vacations, including trains, rentals, leisure activities”, all this while taking into account the particular tastes of each person; or “reorganize the five meetings on my calendar so that I don’t work tomorrow.”
Beyond accessing our intimate data, the main characteristic of these AIs will be their ability to negotiate with other AIs in order to optimize the price of plane tickets as well as the meeting dates of employees in the cooperative or competitive logics described by the mathematician John Nash.
EUROPE SUSPENDED FROM AI ACT
Ultimately, it is possible that our civilization will be largely supported by hundreds of millions of AI assistants, constantly negotiating (in a language that would be equally unknown to us) to optimize our uses. Functions that still take up a lot of our time today (organizing vacations, our trips, the family budget, administrative management, our appointments, etc.) could become invisible in a few years. This is not science fiction, very credible small-scale experiments are already being tested.
And Europe? It is largely suspended from the announcement of the AI Act, currently in “trilogue” discussion between Parliament, Commission and Council. To imagine for a second that this text will turn the tables would be fatal. If Europe wants to get back into the race, it will have to make strong political choices on financing, talent attraction, the education system and many other points.