The IST Distinguished lecture “Unlocking generative AI with ubiquitous hardware and open software”, organised by the Department of Electrical and Computer Engineering (DEEC) and the Department of Computer Science and Engineering (DEI), took place on 25 October, at 2.30 p.m., in the North Tower (room EA1) at Técnico – Alameda Campus. The invited speaker was Wei Li, Vice President and General Manager of Artificial Intelligence Engineering Software at Intel.

For about an hour, the researcher argued that “initial design decisions [of generative AI] that emphasise the use of ubiquitous hardware and open software are critical to ensuring wider future access”. Wei Li, who described himself as the leader of “a team of engineering magicians”, seeks to spread the use of AI, began his presentation with the assurance that, compared to more primitive iterations of other eras, “this time, the AI [being discussed] is very real”.

When asked about the biggest challenges facing these technologies at the moment, the doctorate in Computer Science from Cornell University, says “the biggest obstacle at the moment is probably the idea of responsible AI”. “Before it can be disseminated everywhere, we need to make sure that the AI we’re developing is good AI,” says Wei Li. Adding to the fact that, besides being powerful, “AI is much easier to make accessible” and therefore “controlling it is a challenge”.

In this regard, the speaker spoke about the role of institutions like Técnico in solving problems. He argues that since “AI changes very quickly” and that “new ideas often come from universities”, it is necessary to see students as “future leaders of these technological changes – they have both opportunities and future responsibilities to understand what the solutions are”. Wei Li believes that many of these challenges will take years to solve, so “the ‘new blood’ will bring the bright new ideas” that will do it.