Anthropic CEO Dario Amodi speaks out on 'AI regulation,' refutes claims that regulation will lead to centralization.


by

Fortune Brainstorm Tech

Anthropic CEO Dario Amodi has expressed support for the AI regulation law passed in California, sending the message that AI can be dangerous. This has led some to interpret it as him wanting to centralize and monopolize AI, prompting Amodi to explain his intentions.




The whole thing started with a comment made by technology investor Gavin Baker on a podcast . Baker said that he had heard from 'several trusted sources' that 'CEO Amodi said Anthropic might one day become the only private company in the world.'

CEO Amodi wants proper regulation of AI. He supports the bill passed in California to ensure transparency in AI and has also sent a message in favor of government regulation, arguing that the threat of AI will increase if it is not regulated. However, this stance has been seen as leading to the idea that powerful AI should be controlled by only a few companies, and as justifying centralization. Combined with Baker's remarks, some people have begun to think that 'Anthropic itself may end up wielding immense power.' In fact, California has partnered with Anthropic to implement an AI discount program for local governments, so it's not surprising that some people think 'they are trying to create an environment that is advantageous to themselves by regulating AI.'

California partners with Anthropic to make Claude available to state agencies and local governments at half price - GIGAZINE



Sholto Douglas, a researcher at Anthropic, was the first to counter Baker's remarks, pointing out that 'Baker's statements are lies tailored to the narrative that certain people believe, and Anthropic is most concerned with the concentration of power.' Baker responded that the problem is that many people actually perceive Anthropic as wanting to centralize powerful AI due to its perceived dangers. He suggested that 'there will be two options: either concentrate AI in the hands of a select few companies and politicians through regulation, or distribute it widely,' adding that he personally believes the latter is more important for the development of AI.

CEO Amodi himself joined the exchange and shared his intentions.

CEO Amodi stated that while he supports AI regulation, he has always maintained a cautious stance. For example, he explained that he supported the aforementioned California bill because it exempts low-profit companies from regulation, which aligns with Anthropic's proposal to 'regulate cutting-edge AI companies and favor smaller competitors.'

Next, regarding the idea that 'regulation leads to the concentration of power in the hands of a few companies,' he countered that it was 'an overly simplistic worldview.' He stated that regulations are generally thought to limit the power of companies and benefit the general public, and that it largely depends on what 'regulation' is made of. In addition, CEO Amodi emphasized that 'it is important to set the right rules,' and said he supports an approach that addresses the risks of AI, limits the power of cutting-edge AI companies as needed, and requires pre-screening before AI is made publicly available.


by Fortune Brainstorm Tech

Regarding the message that warned about the dangers of AI, he countered, 'I see opinions that I was exaggerating the dangers, but in reality, I was expressing my opinion on the risks and benefits of AI to an equal degree.' He shared the story of his father dying of hepatitis C just before a cure was completed, and conveyed that 'AI also has the ability to develop new drugs, and I am suggesting that this ability not be hindered by regulations.'

He also acknowledged that many people are skeptical of clichés like 'AI will cure cancer,' adding, 'It certainly deserves criticism, and the only way to build trust in AI is by actually curing cancer. That is our responsibility.'

CEO Amodi said, 'Anthropic is making very rapid progress in the fields of biology and medicine, and I believe we will achieve incredible results in the coming years, and we will be able to make some early announcements in the coming months. When we actually accomplish something, I promise to announce it as loudly as possible. But until then, I don't want to make empty promises. On the other hand, I am being honest about the real risks that AI brings and how we are dealing with them.'

Related Posts:

in AI, Posted by log1p_kr