Suleyman stressed the importance of regulating AI to prevent misuse of information and its impact on nation states. Policymakers and regulators across the world must think of regulating artificial intelligence (AI) in the next 5-10 years to prevent misuse of such tools in affecting nation states by politicizing the spread of information, said Mustafa Suleyman, cofounder, Inflection AI and DeepMind.
“I think the negative consequences here are likely to arise simply because so many people with different organisational interests, different religious and political values will not only affect information but also the real world in practice. I think that is the challenge for nation states,” Suleyman said.
“That’s where in the next 5-10 years regulation will be required to eliminate what can be done using these resource tools,” he said in a virtual chat at the Nasscom Technology and Leadership Forum 2024 on Tuesday.
The Silicon Valley thought leader has been vocal about explicitly and promptly banning the use of AI and chatbots in electioneering.
In his book, The Coming Wave, Suleyman has argued that all technology is political, which can undermine the information space with deepfakes and targeted, adaptive misinformation that can emotionally manipulate and convince even savvy voters and consumers.
He added that risks like cyberattacks, chemical and biological weapon development, proliferation of child sexual material, can still be managed, because people building AI are responsible citizens with established governance and organisational controls.
Suleyman is also wary about increasing self-consciousness of AI which, in future, may claim that it has feelings and emotions. “This is going to be problem for AI ethicists.”
He added that engineers at Inflection AI have exercised these controls while building the Pi chatbot.
“If you ask Pi how are you feeling, it will refrain from answering, it is reluctant to show self-awareness,” Suleyman said.
The Palo Alto, California based startup, last year, debuted with the Pi chatbot, which stands for ‘personal intelligence’, focusing on the needs of a single human user, instead of tasks such as computer coding enabled by the likes of ChatGPT.
“Everyone is going to be represented by AI. You are going to have AI representing your interests on your behalf, presenting your brand value, your ideas…it will be your interlocutor,” Suleyman said during the Nasscom event.
“We are already seeing people having romantic relationships with AI, already seeing doppelgangers of celebrities.”
Inflection AI is currently valued at $4 billion after raising $1.3 billion last year by Silicon Valley heavyweights including Microsoft and Nvidia. LinkedIn co-founder Reid Hoffman, Bill Gates, and former Google CEO Eric Schmidt, also participated in Inflection’s funding round.