Big Data & AI Exhibition, Paris – At a time when AI has become a vast marketing trap, he is one of the only ones to demystify it unvarnished. “AI has nothing to do with intelligence” says Luc Julia, Scientific Director of Renault, and veteran of Silicon Valley.
So why do we keep saying that machines are intelligent? The error comes from 1956, when we had to give a name to a technology. “It’s machine learning, deep learning, but not intelligence,” insists the man who set up Siri.
The fact remains that with generative AI, which has invaded the field of IT since the success of OpenAI, we must find a definition. “What’s interesting here too is the term “generative” says Luc Julia. These AIs generate, but they don’t create anything. Creativity remains on our side.”
Gen AI “exacerbates your creativity”
And to illustrate the point. “I’m bad at drawing, he says. But with gen AI I create, and they generate, with Stable Diffusion for example. They generate, and I repeat the requests until I find exactly what I want as a result “. He therefore deduces that Gen AI “exacerbates your creativity”, and that they are therefore not likely to take your job.
Luc Julia creates green dogs on the Eiffel Tower. And he seems happy with the result.
But GenAI does contain a small revolution. “AI is an evolution, with more computing, more data since 1956. But with ChatGPT, the revolution is in the chat part. It becomes very easy to ask anything.”
“If Gartner has just put generative AI at the top of its hype cycle, it’s because of the 100 million ChatGPT users” says Luc Julia.
“If Gartner has just put generative AI at the top of its hype cycle, it’s because of ChatGPT’s 100 million users. But where does this incredible popularity come from? Well, it comes from the ease of use the cat” he said.
Renault design has embraced generative AI
However, he predicts that the innovation hype is already fading, and that disenchantment will arrive very soon. Then, and only then, will the adoption plateau arrive with applications in the professional world.
At Renault, generative AI found its first followers in design, known as Luc Julia. “Design teams adopted Midjourney much more quickly than Photoshop did 20 years ago,” he assures. Another example of use by the French manufacturer is the use of generative AI for contracts, translation, or manual writing.
“We have seen prompt companies created in recent months in Silicon Valley. But these companies are dying quickly because this know-how is increasingly integrated into the professions” he says to ensure the speed of the corporate adoption.
Move towards fine tuning to gain relevance
The fact remains that this technology must be handled with a grain of salt. “Hallucinations and errors are dangerous. I asked Google Bard for my bio, and everything was wrong,” he says, citing a study on ChatGPT which established its level of relevance at… 64%.
Google Bard does not yet have a very good grasp of Luc Julia’s CV.
And to give some ideas for improving the results of generative AI. Do fine tuning by entering your own data into generative AI to gain relevance, work with open source tools, move towards frugal models to limit energy bills.
Finally, on the question of AI taking control of our lives and our jobs, he is categorical: “We are moving towards specialized AI models. AI will never look like us”.