TIL about homomorphic encryption

It is really easy for me to fall into the trap of letting an AI agent do work *for me* instead of using it as a thought partner to *help me*. I think this is a common pattern for folks. The former is the path of least resistance. The latter leads to better, more valuable output, but requires me to actually think.

TIL Brandolini's law (aka the bullshit asymmetry principle)