hello... I have started a small Zulip, on which we are reading Munkres' "Topology". in weeks 1 and 2 we have just been reading about sets and stuff (Ch.1). but on Monday we are going to learn what a "topology" is (Ch.2). so, still a great time to join https://t.co/s7dBYWK0B8
my multitalented wife @holotopian is looking for a job after a couple years off. she's looking for analyst / ops-y / coordination-esque / weird roles - she's incredibly talented at managing complex projects and getting shit done, and during her break has gotten familiar with different agent tools (learning from the best 🙂) so knows her way around both a RACI chart and a claude swarm. her portfolio of projects is at https://t.co/2ftNdoKuLX.
as an example of her analyst-coordinator prowess, in her last job she took charge of a smartphone migration at a children's hospital that two previous hires had failed to fully tame. she munged (with python, pre-llms!), merged, and analyzed multiple incompatible device lists to get a master list of 6,000 missing and out-of-date phones and tablets spread across different doctors, labs, and departments at the hospital, got them all correctly categorized, upgraded, and entered into the system they were supposed to be in, and also saved the hospital $17,000 a year on phone plans in the process.
she's looking for something different, on the small side, full or part time, remote or seattle-based, maybe overlapping with biology (one of her degrees) or japanese (her other, she's passed the N2.) and i'm biased but she's really cool. you should hire or refer her!
@voooooogel basically to me (as someone who knows hardly anything about this stuff tbh) "ouch" and "unction" share the vibe of being very common word-suffixes but rare words (but not nonexistent words!)
@voooooogel maybe those tokens are more highly-predicted than other bad tokens just because they are popular?
there are a lot of other words ending in -ouch and -unction, like slouch, or conjunction... how are those words tokenized?