Fei-Fei Li, the Stanford researcher who additionally based World Labs, was invited to the Grand Palais in Paris on Monday to provide the opening speech on the Synthetic Intelligence Motion Summit which is going down this week.
Whereas she shared a few of her speaking factors over the weekend, her stance on AI regulation and the AI ecosystem as an entire was fascinating. She managed to seem each pro-business and anti-Massive Tech/massive AI labs on the similar time.
“We have to spend money on far more healthy and extra vibrant AI ecosystems the place academia and the neighborhood can produce,” she informed convention delegates. “Open-source communities and the general public sector can all take part and play their essential position alongside massive firms in driving this expertise ahead. If AI goes to alter the world, we want everybody from all walks of life to have a task in shaping this alteration.”
“If these sources are concentrated in solely a handful of firms, the AI ecosystem will undergo from the dearth of curiosity-driven analysis, high instructional skills, open-source efforts, and multidisciplinary explorations,” Fei-Fei added.
On the similar time, she shared her considerations about AI regulation world wide. She didn’t single out any authorities or legal guidelines specifically — however known as for a extra pragmatic strategy.
“For starters, it’s important that we govern on the premise of science, not science fiction. A lot of right now’s AI conversations are lined by sensationalism and lead to deceptive insurance policies of AI governance. As a substitute, we have to apply a way more scientific technique in assessing and measuring AI’s functionality and limitations, which might consensually result in extra exact, actionable insurance policies round it in actuality.”
She additionally warned towards pointless restrictions on AI analysis.
“In flip, this brings me to the second level of embracing a realistic perspective as a substitute of an ideological one relating to AI governance. AI is poised to be a strong expertise that may assist us to dwell and work higher if used effectively,” she mentioned.
“So as a substitute of hindering the upstream exploration of analysis of this nonetheless very younger expertise, we should always put extra deal with the sensible purposes in order that we will guarantee its benevolent utilization to protect towards dangerous outcomes,” she added.
In accordance with Fei-Fei, historians will look again at this era in time and name it the primary true period of AI, which may have “civilizational penalties.”
However whereas she acknowledged that jobs are more likely to be impacted by the progress of AI, she additionally mentioned she believes that AI has the potential to “increase and empower us” — assuming, that’s, policymakers, AI firms, and the analysis neighborhood make the suitable choices.
Authorities leaders and tech CEOs are anticipated to signal and share an announcement on the finish of the AI Motion Summit on Tuesday.