Ex-OpenAI CTO says humanity still has agency over frontier AI models

The artificial intelligence landscape is contending with mounting concern over the unprecedented self-evolving capabilities now being demonstrated by frontier AI models. Anthropic has emerged as the most vocal voice amongst major technology companies, calling on the international community and the wider tech industry to consider a freeze on the development of advanced AI systems.

In a recent blog post, Anthropic issued a stark warning, asserting that AI models are approaching the capability to improve themselves without any human intervention. The company's concerns are particularly acute given that its own model, Claude, has reportedly demonstrated the ability to improve AI code up to 52 times faster than a human — and may soon be capable of building its own successor.

Murati offers a different perspective

Amid the alarm, former OpenAI Chief Technology Officer Mira Murati has offered a notably more measured outlook. Speaking to Emily Chang at Bloomberg Tech, the CEO of Thinking Machines Lab argued that the outcome of frontier AI development is far from predestined, directly countering the wave of pessimism that has come to dominate conversations about the technology's future.

"Predicting a dystopia or a utopia, to me, feels very simplified because the truth is we actually have a lot of agency in how we build this technology, in the tools that we're building, how we're deploying it," she said.

Acknowledging the risks whilst emphasising human agency

Murati was careful not to dismiss the very real risks that accompany the development of powerful AI systems. "We all understand the potential for greatness that comes with building frontier AI systems, and that's why we're working on them," she acknowledged.

Rather than viewing the situation as one spiralling beyond human control, Murati placed considerable emphasis on the agency that humans retain. She described the present moment as a unique opportunity, in which humans and AI are navigating the path forward together — and argued that this is precisely the right time to shape that partnership through meaningful collaboration.

"The more we can reduce discontinuity that comes from new capabilities, the better," she added.