Vitalik Buterin stresses AI risks amid OpenAI leadership upheaval

Vitalik Buterin calls superintelligent AI "risky" amid leadership upheavals at OpenAI, stressing the need for caution and decentralization in AI development.

The Ethereum co-founder Vitalik Butertin has shared his take on “superintelligent” artificial intelligence (AI) calling it “risky” in response to ongoing leadership changes at OpenAI. 

On May 19, Cointelegraph reported that OpenAI’s former head of alignment, Jan Leike, resigned after saying he had reached a “breaking point” with management on the company’s core priorities.

Leike alleged that “safety culture and processes have taken a backseat to shiny products” at OpenAI, with many pointing towards developments around artificial general intelligence (AGI).

AGI is anticipated to be a type of AI equal to or surpassing human cognitive capabilities—the thought of which has already begun to worry industry experts, who say the world isn’t properly equipped to manage such superintelligent AI systems.

This sentiment seems to align with Buterin’s views. In a post on X, he shared his current thoughts on the topic, emphasizing that we should not rush into action or push back against those who try.

Buterin stressed open models that run on consumer hardware as a “hedge” against a future where a small conglomerate of companies would be able to then read and mediate most human thought. 

“Such models are also much lower in terms of doom risk than both corporate megalomania and militaries.”

This is his second comment in the last week on AI and its increasing capabilities.

On May 16, he argued that OpenAI’s GPT-4 model has already exceeded the Turing test, which determines the “humanness” of an AI model. He cited new research that claims most humans can’t decipher when talking to a machine.

Related: Microsoft’s new ‘Black Mirror’ recall feature records everything you do

However, Buterin is not the first to express this concern. The United Kingdom government has also recently scrutinized Big Tech’s increasing involvement in the AI sector, raising issues related to competition and market dominance.

Groups like 6079 AI are already emerging across the internet, advocating for decentralized AI to ensure it remains more democratized and not dominated by Big Tech.

This follows the departure of another senior member of OpenAI’s leadership team on May 14, when Ilya Sutskever, co-founder and chief scientist, announced his resignation.

Sutskever did not mention any concerns about AGI. However, in his post on X, he expressed confidence that OpenAI will develop an AGI that is “safe and beneficial.”