Axios AI+ SF Summit: Experts identify the gaps in AI optimism and opportunity
Add Axios as your preferred source to
see more of our stories on Google.
Sarah Bird, Microsoft chief product officer of responsible AI, joins Axios' Ina Fried on stage at the AI+ SF Summit. Photo: Chris Constantine on behalf of Axios.
SAN FRANCISCO — Policymakers, executives and key technology entrepreneurs at the Axios AI Summit on Dec. 4 see AI's potential benefits but have concerns about large language models, cybersecurity and the tech becoming dangerously autonomous.
Why it matters: With the rapid, cross-industrial growth of artificial intelligence comes the potential for risk. Key players in the space hope to harness its potential without jeopardizing the wellbeing of industries and social good.
- The event was sponsored by AstraZeneca, Atlassian, Bank of America, Meta and Securus Technologies (an Aventiv company).
The five key takeaways:
- LLMs have a long way to go, says Geometric.ai founder Gary Marcus. Though Box CEO Aaron Levie is more bullish, he agrees that "the technology has limitations. It fundamentally requires humans to provide the oversight."
- New AI-powered cyberthreats make innovative detection strategies a priority, says Palo Alto Networks CEO Nikesh Arora, and the nature of AI's reliance on personal data makes those systems particularly tempting to attackers.
- Transformation is on the horizon, but there are dangers, according to Google DeepMind CEO Demis Hassabis. Among them: a crisis of human purpose, bad actors weaponizing AI technology for harm, and the AI itself becoming more autonomous and going "off the rails in some way that harms humanity."
- The public wants AI it can trust, said Sarah Bird, Microsoft's chief product officer of responsible AI. "I've been very positively surprised with generative AI, how much the world has woken up and said, 'Responsible AI is incredibly important and we don't want this technology without it,'" she said. "And we've seen the regulations starting to follow with that. We see customers demanding it."
- AI remains important to those outside of the tech bubble, though it isn't always as urgent as "housing and cost of health care and secret police grabbing your neighbors and sending them to gulags in El Salvador," said California state Sen. Scott Wiener to Axios' Ashley Gold.
Content from the sponsors' remarks:
Atlassian chief marketing officer Zeynep Inanoglu Ozdemir said that "the lowest hanging fruit in how AI can unlock unprecedented pace of learning is bringing your data together, and making it available — searchable, discoverable, indexable — and having conversations with it."
Michelle Boston, Bank of America CIO and head of data management technology & enterprise architecture, said, "The data and the AI are two sides of the same coin." Anyone using AI must start with a strong foundation of high-quality, accurate and complete data — including guiding the technology with context.
Go deeper: Watch the full interviews on YouTube.
