I don’t follow Mandy Patinkin in Instagram, but enough of my followers and people I follow, follow him that his post from this morning showed up on my feed. It was a series of screenshots from an X.com user named Jacob Coxon who says that he resigned from Anthropic yesterday. I read the rest of his post, which was quite alarming, actually terrifying. Here’s a scary nugget…
Like all things on social media, I wanted to confirm Coxon was a real person and truly used to work at Anthropic. Sure enough, he is very real, and Variety wrote a piece about his X post here. They reached out to Anthropic and got ahold of two senior-level employees who basically affirmed what Coxon said in his X post. From the article: Hubinger wrote the company did “earnestly believe AI could kill all humans” and that he thought it was a greater than 10% risk “within the next desk.”
The Variety article is an excellent summary of Coxon’s post, Anthropic’s response, and other relevant quotes from very powerful people explaining why companies are racing toward a super intelligence they will not be able to control and have already lost understanding of.
We have come to expect companies to compete against each other, so it is no surprise that Anthropic and OpenAI (and others) aren’t slowing down even when it seems the end goal will literally mean our end. And that’s not concerning to our government either. Just yesterday, the Treasury Secretary Scott Bessent was quoted, “There is no day after tomorrow if China wins at this…If they were to pull ahead of us on AI, then nothing else matters.” What Bessent clearly doesn’t grasp is that it doesn’t matter who wins the race to an AI super intelligence. Can someone explain to him that there is not a trophy or a ribbon or a big bottle of champagne at the end of this race? Even the people who are in it know there’s already a 10% chance (they’re being conservative) that the participation prize at the end of the AI race is some super intelligent, sentient AI telling us, “Congrats, humans. Now you don’t get to live anymore.”
Furthermore, Bessent is assuming that if we get the technology first, we will use it for the good of the world, we will use it responsibly, and that any other country will use it to our harm, right away, and intentionally. That hasn’t been the case with the United States and deadly technology in the past. We are super good at turning anything into a weapon, but the next weapon will just turn on us. It will be too late at that point.
Leave a Reply