Big Tech Didn't Write the AI Ban Bill That Would Tank Its Own IPO
Malone News says Silicon Valley engineered an AI safety panic to bury smaller rivals in compliance costs. The bill it blames would gut the valuation of the company it names, and the group actually spending the most money is fighting the opposite fight.
Am I The Idiot — the strongest case against our own take.
The strongest case against this piece is that it may be answering a narrower question than the one that matters. I argue a full ban on superintelligent AI would hurt Anthropic's valuation, therefore Anthropic is not the beneficiary of that specific bill. But a company does not need a bill to pass, or even to pose a genuine threat to itself, for that bill to be politically useful to it. A hard, plainly extreme ban bill can function as a radical flank: it makes Anthropic's own preferred policy, a disclosure-and-testing regime enforced by embedded evaluators, look moderate by comparison, precisely because the Sanders-Casar version sounds unworkable. Amodei's own essay, published nine days after the bill and four days after Coxon's post, proposes exactly that comparatively modest alternative. If the ban bill's real function is to shift what counts as "reasonable" AI policy rather than to actually pass, it can serve Anthropic's interests even while being bad for Anthropic's stock price if enacted, and this piece's framing understates that possibility.
I should also be fair to the coordination question I largely dismiss. I found no evidence that Coxon's resignation was choreographed, and every outlet treats it as organic. But an absence of evidence in under two weeks of reporting on a fast-moving story is a weak instrument for ruling coordination out entirely; it mainly shows no journalist has proven it, not that it did not happen. A more cautious version of this piece would simply call the coordination claim unproven, rather than leaning, as this piece does, toward treating it as unlikely.
Finally, attacking Sayer Ji's credibility does not touch whether the dollar figures in the original piece were sized correctly. If Ji's numbers happen to match the independently sourced totals found here, the sourcing complaint is a matter of professionalism, not accuracy, and a reader who only cares about the second may reasonably find this piece's emphasis on the first unfair.