The AI progresses quickly but arouses many concerns, even with Pope Francis. Questioned by TechCrunchresearchers say they are exhausted by this pressure in the face of rapid progression in the sector. One of them confides: “Everything has changed overnight, with an enormous impact of our work, positive as negative, measured in terms of exposure of products and financial consequences.”
Advertising, your content continues below
The rapid progression of AI is problematic
The industry announces novelties at a frantic pace. OPENAI, which launches its Stargate project at $ 500 billion and will add an additional layer of stress, organized 12 conferences in December 2024. The company of Sam Altman announced more than a dozen new tools, models and services.
Ditto for Google with an avalanche of press releases with posts on social networks and blog notes. In short: the cadence is hellish and has a considerable human cost.
It is common for IA researchers to work six days a week at Openai with exhausting hours. Sam Altman pushes his teams to transform their discoveries into public products with tight deadlines. Bob McGrew, former research director, cites professional exhaustion as one of the reasons for his departure last September.
It is hardly better in other laboratories. The Google Deepmind team to whom we owe Gemini went from 100 to 120 hours of work per week to correct a bug. XAI engineers, Elon Musk’s company, often work on the first morning lights.
Pressure is also explained by the considerable financial impact of AI research. Gemini’s bug has lost $ 90 billion in market value in Alphabet. Kai Arulkumaran, research manager at Araba, specifies that “One of the greatest pressures comes from competitiveness, combined with very short deadlines.”
Advertising, your content continues below
-Competition plays on rankings like the Arena chatbot which assesses AI models. Logan Kilpatrick, product manager at Google Gemini, recognizes the impact “Not negligible” of these rankings on the development speed of AI.
What solutions in front of this sharp pace?
The frantic race therefore worries the researchers who see their work exposed to obsolescence even before a publication. Zihan Wang, robotics engineer in a startup, wonders: “If the probability that someone goes faster than me is huge, what is the meaning of my work?”
Gowthami Somepalli, doctoral student at the University of Maryland, shares this difficulty in following the rhythm of publications, in particular to distinguish real advances from fashion effects. After two years of thesis, the researcher ceased to take a vacation by guilt not to publish anything.
However, solutions emerge in the face of this situation. Bhaskar Bhatt, consultant at EY, plans a support network between researchers. Ofir Press, post-doctorant in Princeton, suggests reducing the number of conferences and setting up weekly breaks in the publication of articles. Raj Dabre, researcher at NICT in Japan, recalls that it is necessary “Educate people from the start that AI is just a job” And focus on family, friends and more essential aspects of life.
Advertising, your content continues below
Tech
Related News :