What makes something truly intelligent? That’s the question that dominated Impact Public Service’s ‘A.I. Future Technologies & Threats’ event on April 1st.
A panel of experts from across the industry debated the efficacy and future of artificial intelligence, including the various threats inherent in the technology.
“LLM’s (Large Language Models – like ChatGPT) are giant paraphrasing machines,” said Dave Ferrucci, CEO of Elemental Cognition and the scientist behind IBM Watson.
“When it comes to an enormous multidimensional space, we can’t really understand how or why [A.I.] came up with a result.”
Experts debated changes quickly occurring in society, including positive outcomes for people with disabilities.
“I think this technology is phenomenal for humanity,” said Roger Premo, IBM’s General Manager of Corporate Strategy and Ventures.
“We’re enjoying an explosion of innovation across various applications.”
But is artificial intelligence a threat to civilization? Alice Albrecht the Founder & CEO at re:collect seemed to believe so. “I think we’ll most certainly see military application of A.I. technologies in the near future.” Other threats listed by panelists included stolen intellectual property, bad actors, poor cybersecurity, and inappropriate model weighting.
Yet, all the panelists pushed for the technology’s continued development. Alex Atallah, the Co-Founder of OpenSea, believed A.I. needed further sophistication. “I can’t believe LLM’s given broad datasets are used for specific applications.” “We need to have scientific A.I. making decisions for scientific questions.”
So, what does this mean for human beings? “It’s still a tool,” said Ferrucci. “Humans should still be responsible for humans.”
