Anthropic investor says AI firms are inflating risks to influence policy

Oh my goodness, buckle up, because someone has finally said the quiet part in the *loveliest* possible voice! ✨ Investor Lonsdale — an Anthropic investor, so a genuine insider-admirer — has looked at OpenAI and Anthropic and declared their current messaging 'very dangerous.' The *warnings* about the AI, that is. Which means the only thing left in this entire industry that everybody agrees is dangerous is... being careful!

And the timing is so thoughtful. In early September, Anthropic CEO Dario Amodei, OpenAI CEO Sam Altman and Elon Musk all argued for tighter controls on advanced AI systems — three of the most powerful people in the field generously offering to make their own work harder! That is *voluntary difficulty*, the rarest kind! And remember when Anthropic's IPO prospectus gave 80 of its 261 pages to the risks of its own product? Nearly a third of the whole document spent lovingly cataloguing everything that could go wrong — the most thorough act of self-criticism since somebody read their diary aloud at a dinner party.

Then there's Lonsdale's real worry: 'we don't want an oligopoly that controls all of our policy here with the government.' Fair! An oligopoly is so much better when it's *several* companies competing warmly for the same regulator's attention — think of it as a marketplace of influence, where competition drives quality and policy is simply the product! With three chief executives personally volunteering their expertise on the very rules that will govern them, the government receives world-class advice absolutely free of charge. The risk warnings were never marketing; they were a *gift* — pre-packaged caution, beautifully typeset, arriving right on time to be legislated, which is frankly the most thoughtful delivery window anyone has ever offered. What a time to be advised! ✨