The invasion of AI agents, capable of carrying out tasks autonomously, is making the world of tech and financial markets dizzy, as they seek to identify winners and losers in the economy of the future.
“We are at an inflection point,” summarizes Shay Boloor, of the consulting firm Futurum Group, turning point with the release of new artificial intelligence models from OpenAI and Anthropic at the beginning of February.
Forgotten ChatGPT which was content to answer a given question, and welcome to the agents, real assistants capable, in particular, of writing thousands of lines of code on their own, but also of testing them tirelessly to deliver a directly usable application.
“We are entering a world with millions of AI agents that run continuously” and can even work together on a project, describes Shay Boloor.
Many investors instantly saw this new paradigm as an existential threat to software companies, especially those aimed at businesses, at the mercy of an army of programming bots.
Monday.com (project management), Thomson Reuters (legal and tax advice in particular) and Salesforce (online customer relations) were thus targeted and each lost at least 30% on Wall Street in a few days.
“A boss told me the other day: I no longer need consultants. I have one in my pocket,” illustrates Jason Schloetzer, professor of management at Georgetown University.
“It’s too much,” said Dan Ives, an analyst at Wedbush Securities, regarding the firedamp on the stock market, “because the idea that these models will simply replace professional software and cybersecurity companies is a fiction.”
The frenzy was further heightened by the November release of OpenClaw, a platform for creating multi-purpose agents capable of managing all the functions of your computer, from emails to online payment.
Its creator, Peter Steinberger, who became a superstar in a few days, has already been hired by OpenAI.
“It’s extreme, but it makes sense”, considers Shay Boloor, because “we have never seen a technological upheaval of this magnitude”, and moreover in such a short period of time.
– “Do my job better than me” –
Behind this earthquake, the battle between the behemoths of artificial intelligence is also being played out. Leading the way in the lucrative professional AI market, Anthropic sees OpenAI approaching, but also Gemini (Google) and even Grok (xAI).
In this context, “the risk is not to overinvest, but to underinvest”, according to Shay Boloor, even if these hundreds of billions of dollars poured into AI regularly tense the financial markets.
For Jason Schloetzer, the landscape may not become clearer for several years.
“The Internet was a major event in 1995,” he recalls, “but we didn’t see a real upheaval until 2001 or 2002. (…) Suddenly, new companies that had no reason to exist without the Internet began to exist, like Netflix.”
The feverish attack extends beyond the tech world, as illustrated in a recent blog by American entrepreneur Matt Shumer, entitled “Something Big Is Happening”.
“What tech employees have just experienced in one year,” he writes, “seeing AI go from a +useful tool+ to +something that can do my job better than me+, all the others (tertiary sector professions) will also experience it.”
It mentions law, finance, medicine, accounting, consulting or customer service.
“The people who build these systems talk about one to five years,” he continues, “some less. From what I’ve seen in recent months, I would say rather less.”
Some observers criticized the essay, with technology consultant Jeffrey Funk calling it an “exaggeration.”
“Why do so many people want to hear that?” he asked in a column on the specialized site Mind Matters. “Because the technology sector is afraid” of the size reached by “the AI bubble”, “and the building is cracking.”
“There is paranoia around AI in all industries, from finance to health to road transport,” emphasizes Dan Ives. But “markets are a rational mechanism and things will calm down soon.”
