This compilation technique brings very interesting results. Hopefully should find its way in some JIT compilers.

Maybe it's time to make so called "reinforcement learning from human feedback" actually humane? It's not the first account along those lines in the industry.

Good set of advises for Python APIs. Some applies more generally though.

We went from quality to quantity it seems. We also have whole swats of developers who are just consuming content without critical thinking and it's a problem. The conclusion says it all: "Don’t consume. Create. Ask questions. Stay curious."