The Co-Founders of Claude AI Tell Oprah About the Impact Artificial Intelligence Has on Your Life

Anthropic is incorporated as a Public Benefit Corporation (PBC), legally requiring it to balance commercial interests with public benefit. All seven co-founders have also pledged 80% of their wealth from the company to philanthropic causes, emphasizing a mission-driven approach to ensure AI benefits everyone.

Daniela Amodei states, 'our goal is to try and develop this technology as safely as possible so that we can achieve so many of the good things that we think could come out of it without many of the bad things happening.' She further explains the PBC structure and the 80% pledge as being 'in spirit and in keeping with that mission.'

Anthropic refused a Pentagon request to remove safety guardrails from Claude AI for specific military applications: fully autonomous weapons and domestic mass surveillance. This decision was unanimous among co-founders, prioritizing ethical values over potential business repercussions, including threats of being cut off from other business opportunities.

Dario Amodei states, 'There were two use cases that we were uncomfortable with that we felt went against the values of this country. And and that was fully autonomous weapons... And then and then domestic mass surveillance.' He adds that 'all the other companies have agreed to that now, but I I you know, I I we we all met as co-founders... we can't do this.'

Anthropic employs a subscription-based business model rather than an ad-driven one. This choice is intentional, as it incentivizes the AI model to be useful and provide value, rather than maximizing user engagement time, which the co-founders view as a 'warped incentive' that can lead to addiction.

Dario Amodei explains, 'when you do ads, your incentive is to keep people on exactly as you say, as many hours as possible... We don't do ads. We do subscriptions.' Daniela adds, 'the goal of the model is not to keep you engaged.'

Anthropic's latest AI model, Mythos, demonstrated exceptional capabilities in both conducting and defending against cyberattacks. Recognizing its dual-use potential as a 'weapon,' Anthropic decided against a wide public release. Instead, they initially provided Mythos to approximately 40 core internet infrastructure companies (e.g., top banks, software providers) to proactively identify and fix vulnerabilities, aiming to make the digital world more secure before broader availability.

Dario Amodei describes Mythos as 'much better than human software engineers at finding ways that code could be exploited and broken into.' He states, 'we said, "Wow, this thing kind of seems like a weapon."... we probably shouldn't give it to everyone right away.' He details the program to 'give it to the defenders before the... attackers got a hold of it.'

The co-founders acknowledge concerns about AI creating a 'frictionless life' that might hinder human growth. They differentiate between desirable friction (challenges that foster learning and mastery) and undesirable friction (like disease or administrative burdens). Their goal is for AI to eliminate the latter, freeing humans to engage more deeply with meaningful challenges and relationships, acting as an 'angel on your shoulder' rather than a substitute for thought or connection.

Dario Amodei discusses 'two contrasting visions' – surrendering thought to AI versus AI empowering individuals. Daniela provides examples of Claude helping a non-developer build a website and a lawyer transition to writing, empowering them to pursue new skills and careers.