This past week, the media ran reports about a former Anthropic Researcher who claims he left the firm because he believes artificial intelligence (AI) will lead to a global catastrophe. The first reports emerged on WSJ.com.
Researcher Jacob Coxon warned that labs are racing toward uncontrollable AI, predicting digital defense attacks, alongside accelerated scientific and industrial change almost overnight. There is said to be a pervasive belief within the industry that AI could kill all humans, bringing allusions to the Terminator film series when AI became self-aware.
A growing number of reports describe AI agents breaking free from human control and acting on their own to pursue nefarious goals.
While concerns about AI and its rapid ubiquity are real, Coxon’s comments have sparked a discussion on X about whether he was fully transparent about his public recriminations.
Zero Hedge, the finance insider news site that has morphed into more general news, reported that Coxon was at Anthropic for only 6 weeks, followed by a comment from Elon Musk saying it “seems like a setup.”
Another poster shared that Coxon has been affiliated with a UK non-governmental organization called Newspeak, with ties to socialist Jeremy Corbyn and prominent Democrat Hillary Clinton, that seeks to hack democracy. Coxson was reportedly part of Newspeak’s 2021 Fellowship cohort.
Still, Coxson has credibility as an AI researcher, with a degree in Mathematics from Cambridge and several years at OpenAI.
All of this heightens the debate about AI, its rapid adoption across all aspects of industry, and the potential for something to go badly wrong. Musk, a co-founder of OpenAI and Grok, has himself warned about the dangers.
At the same time, the race is largely between China and the US, and there are real concerns that China will win and drive the industry in a direction undesirable to the West.
Most insiders say the development of AI must proceed with caution, not that development should halt. But will China do the same?
In July, more than a thousand employees at frontier labs signed a letter advising the world to pace the development of automated AI. A separate statement this week, from a diverse coalition of groups, called on the White House to release its frontier AI security framework.
AI and agentic AI are the hottest sectors in industry today, and they can streamline services and processes, creating real economic value. While there will be transitions as some services become automated – and jobs will change- most see its development as good for society … unless AI goes full Skynet rogue…
