“This is a crisis of trust”: Anthropic head explains people’s negative attitudes towards AI

Developers of advanced American artificial intelligence models often issue warnings about their use, and in this regard, Anthropic head Dario Amodei often makes very pessimistic arguments. At the same time, he himself believes that he is only warning humans of risks, and that society’s negative perception of artificial intelligence is largely caused by a “crisis of trust.”

Image source: Human

As Anthropic CEO himself emphasized TechCrunchI acknowledge that this bias exists in society and think it’s a big problem, but I can’t agree with investor Gavin Baker, who accuses the startup’s management of giving others an overly pessimistic view of artificial intelligence. “I think fundamentally it’s a crisis of trust. I think ordinary people don’t trust companies or authorities or the tech industry and always think we’re concocting some new way to deceive them.” – Amodai explains. He added that this cautious attitude in society has been developing for decades and is only just emerging when it comes to artificial intelligence.

The only point Anthropic’s CEO agrees with critics of the AI ​​industry is that it has failed to deliver on its promise to provide general benefits: “It’s really our fault and I think that’s an area of ​​criticism that needs to be focused on, rather than a discussion of messaging and marketing.”. For example, Amodai said the promise of defeating cancer with the help of artificial intelligence is more like a cliché than an inspiring statement. In this case, the only way to change people’s perception of AI is to actually demonstrate its ability to treat cancer.

Amodai also disagrees with liberal experts’ views on authorities regulating the process of releasing artificial intelligence models. This depends not only on the concentration of power in the hands of those who control the release of AI models. “Many people outside the bubble believe that regulation limits the power of corporations and benefits ordinary people,” – He objected. Anthropic seeks to frame its proposals in the regulatory realm of the artificial intelligence industry to curb the expansion of large, advanced companies in favor of the advancement of smaller competitors.

Amodai firmly believes that artificial intelligence is a technology that leads to the concentration of power. The proliferation of open-weight models partially solves this problem, but cannot be viewed as a panacea, as the concentration of power is shifting toward those with the greatest computing resources. He said the industry must set “road rules” that take into account security risks, limit the influence of large companies and leave market space for open-scale models, while taking into account associated risks.