Trump dismissed the possibility that artificial intelligence could wipe out humanity. More than a dozen researchers from OpenAI and Anthropic disagree. They signed letters calling for a slowdown in AI development because they work on the actual models and apparently think those models might kill everyone.
Trump's position makes sense if you consider he's never used a computer. The researchers' position makes sense if you consider they've seen the code and decided their stock options aren't worth planetary extinction. Lawmakers are proposing safeguards, which is Washington-speak for regulations they'll write after consulting with lobbyists who get paid by the companies building the extinction machines.
The irony is perfect. The people building the technology that could end civilization are begging everyone to slow down. The guy who thinks windmills cause cancer says we're fine. One group has PhDs in machine learning. The other group has a guy who stared directly at a solar eclipse.
Retail traders are buying AI stocks either way. They don't care if the models achieve sentience. They don't care if the models decide humans are inefficient. They saw Nvidia go up and they want in. Some guy in Nebraska just took out a home equity line to buy calls on a company that makes chips for data centers, and when the robots take over, they'll use his house as a charging station.
The researchers wrote that rapid AI advances pose serious risks. Trump said he's not concerned. Both statements are true and both are useless. The models will keep getting bigger because no country wants to fall behind and no company wants to miss earnings. The choice isn't between safety and progress. It's between admitting we have no f*cking clue what we're building and pretending we do until something goes wrong. We picked the second one.
Photo by on Unsplash
Leave a Comment