From Stephen Wolfram:
We can say: “Look, this particular [neural] net [works]”—and immediately that gives us some sense of “how hard a problem” it is (and, for example, how many neurons or layers might be needed). But at least as of now we don’t have a way to “give a narrative description” of what the network is doing.
In response to a ‘Dear Neighbor’ letter from a local millionaire, I’m setting the record straight. As we approach our annual town meeting, he is using misinformation to pressure the town into action. My open letter challenges his claims and highlights the truth.
[Post updated on April 11, 2025] In January, I tried to track Trump’s executive actions, but a handy website beat me to it. But with a new authoritarian at the helm, banal summaries don’t cut it. So now I’ve set up a ChatGPT task to do the heavy lifting: provide a brief analysis of each week’s presidential orders from a radical left perspective. No blind trust, no propaganda—just a bullshit detector for the modern age.