On the Rumor of Thinking Machines Sold by Money Men

I have received the strangest dispatch — purporting to come from the year 2026, no less — and I confess it has kept me from my filament tests for the better part of an hour, which is saying a great deal. The claim is this: a concern called Anthropic, building what they term 'artificial intelligence,' is forming a one-and-a-half-billion-dollar joint venture with private equity houses to distribute its product across their holdings. One and a half billion. I have fought tooth and nail for every dollar that built the Pearl Street station, that electrified lower Manhattan, that kept my laboratories running through years when carbonized bamboo was the difference between triumph and bankruptcy. And here these future men speak of such sums as though they were requisition orders for copper wire. Now, let me address the substance. A thinking machine — or something near enough to be sold as one — distributed not through the usual channels of salesmanship but through the leverage of financiers who already own the companies that would buy it. That is not merely clever. That is the strategy of a man who understands that invention without distribution is a curiosity in a glass case. I know this better than anyone alive. My phonograph means nothing if it sits in Menlo Park. My incandescent lamp changes the world only because I built the dynamos, the mains, the meters, the entire system to deliver it into every parlor and office. The product and the network are one. So this Anthropic outfit has learned the lesson. Rather than knock on ten thousand doors, they have found men who already own the doors. I tip my hat to the strategy, even as I doubt the dispatch itself. But I am skeptical of the machine at the center of it. Intelligence is not a commodity like electric current. I have employed hundreds of men in my laboratories, and I can tell you that the thinking part — the real thinking, the part where you try the ten thousandth filament because the nine thousand nine hundred and ninety-ninth was wrong — that is not so easily bottled and sold. If these future engineers have truly built such a thing, then the world has changed more than I can reckon. Still, I notice they need the money men to sell it. Which tells me the machine, however marvelous, does not yet sell itself. And that is a problem I understand intimately.