Large Language Models Are Like The Planet Solaris

It’s been a little while since I watched Andrei Tarkovsky’s Solaris, the Soviet director’s 1973 movie adaptation of a classic sci-fi novel of the same name written by Stanislaw Lem. But the other day, it suddenly struck me that, for being a work from the Cold War, Solaris provides a metaphor that I have been grasping for—a metaphor that helps explain how large language models are able to learn and how they can generate information based on their learnings....

March 22, 2023 · 10 min · 2062 words · Benjamin Levinson

Earthbound in ChatGPT

I just played a surprisingly fun Earthbound-style text adventure in ChatGPT. After this, I’m convinced that we are not very far away from (simple) custom games tailor-made to your tastes. It’s also quite easy to imagine playing an AI-generated Dungeons and Dragons campaign. Prompt: I am playing a text adventure game whose theme is Earthbound, the SNES game. You present me with specific commands to look around my environment and keep track of my inventory....

December 10, 2022 · 1 min · 208 words · Benjamin Levinson

Alfred, or "How to super-charge your Mac and automate routine tasks"

I’ve been reading The Productive Programmer for a book club and it’s put me on a productivity binge. It’s an older book and many of its recommendations no longer apply, so I got excited when, while researching modern alternatives for some of the tools the author recommends, I happened to stumble on Alfred. Alfred is a replacement for the Mac spotlight that has already changed how I work. In fact, I was so excited by Alfred that within ten minutes of downloading, I immediately bought a “mega-supporter” license!...

September 19, 2022 · 3 min · 523 words · Benjamin Levinson