Quick Takes

people generally talk about food preservatives in a negative way. certainly, some of them are not great for you. but I want to take a moment to appreciate how wonderful food preservatives (and refrigeration and pasteurization and canning) are as well. it's crazy how fast most normal food goes bad. like a loaf of real old fashioned bread will go stale after a day and then become moldy after a few more days. for almost all of human history, people just sort of lived with this, and if they wanted to make foods last they had to dry it out and/or drown it in salt or vinegar or alcohol. pickles and beef jerky are great, but it would suck if you had to eat them all the time. 

there seem to be three different possible levels of manager involvement in individual researchers: * type 1 (the grantmaker): the manager spends a day reading the grant application of the researcher and decided whether to fund him for the next 6 months based on his track record, whether the research idea makes sense, etc. day to day, the grantmaker is completely uninvolved in the actual research. on fact, the grantmaker might not even be making decisions about individual researchers, but about entire organizations. * type 2 (the research manager): the manager spends a few hours every month getting complete context on what each of their reports is working on. each report has a different project that the manager thinks makes sense. the manager provides a light touch to guide the project according to their research taste, and vetos any terrible ideas, but most of the time the report makes most of the day to day decisions about what to do. * type 3 (the tech lead): the manager sets research taste for their team and basically orders around their minions. the manager provides almost all of the research taste, and the reports are either just doing engineering work, or doing very tightly scoped research on things the manager has cleared. what I've seen at openai is a mix of all three, but I've found the second one to be the most predominant (my relationship with my managers has been like type 1.75 on average). high level research leaders are generally the first type; they allocate resources to parts of the org they believe to be the most important for the company as a whole. the second type is most people; most researchers have some random direction they believe in, and they are part of a team doing something similar, and their manager has substantial trust in their research taste. even for junior researchers, the focus is on having them develop their own taste eventually, rather than just executing on things for their manager. the third type is most common for large p

I dug a fossil-version Google Glass XE HW2 out of a drawer, pre-rooted, running an AOSP build that I think has a single-digit number of users worldwide. I connected it to the mac mini's USB port, and told my Openclaw instance to get it onto tailscale and set it up as a communication channel. It worked its way through multiple absurd, frustrating technical issues that would absolutely have made me give up if I was the one doing it, with only minimal guidance. Once it had ssh working, it set up an android app. Without me suggesting it do so, it found a way to take screenshots to check its work. So, I have an AI agent on my face now. I don't think it's wise for humanity to be going down this path, especially at this speed, and if humanity gets its act together to pause, I'll power down my Mac Mini and breathe a sigh of relief. But in the meantime, I'm going to enjoy how cool it is, and stay close enough to the forefront to be properly informed, dammit.

Angry ex-wives are the best IRS informants   1. Public criticism of people who aren’t already in the dog house is an act with concentrated costs and distributed benefits. Sometimes those costs are quite high. 2. As such, we should expect it to be underprovided. 3. Occasionally criticism will be provided by very prosocial people, or some other prosocial person will cover some of the costs. 4. But mostly you should expect public criticism to come from people who are being irrational. Maybe they’re generally fragile, or are angry about being dumped for their ex-husband’s secretary. 5. Therefore, you cannot dismiss criticisms because the person is fragile or has a grudge.  6. Under many circumstances- when the target is very well loved and the problem lies in a subtle pattern- the only criticism you should expect will come from people who seem kind of crazy or out to get the target. 7. This is hard to navigate, because those same irrationalities will drive people to make false criticisms too.  8. But "they're irrational/out to get them" is a bad reason to dismiss criticism, because public criticism is an irrational act.

Does anyone have a canonical resource they’d point me to for understanding how to file/optimize (personal, US/California, income) taxes? I found the “Taxes for Dummies” book pretty poorly written; I’m enjoying reading “Taxes Made Simple” by Mike Piper, but it really only covers the basics. Some examples of the sort of thing I’m looking for in other categories: * Get Rich Slowly — great general basic financial advice from putanumonit * Salary Negotiation — great general basic advice on salary negotiation from patio11 * To Do Meetings Well — great general basic advice on how to run & participate in meetings well from Anson Yu * How to Sell — great general thoughtful framing & practical advice on doing “sales” much more effectively/naturally/calibratedly-confidently from Nabeel Qureshi * What They Don’t Teach You in Freelancing School — general advice on freelancing, and more broadly selling your skills/time on the open market from haththerescuer * Butterick’s Practical Typography — general advice on typography

Apparently there's now a sixth person on Anthropic's board. Previously their certificate of incorporation said the board was Dario's seat, Yasmin's seat, and 3 LTBT-controlled seats. I assume they've updated the COI to add more seats. You can pay a delaware registered agent to get you the latest copy of the COI; I don't really have capacity to engage in this discourse now. Regardless, my impression is that the LTBT isn't providing a check on Anthropic; changes in the number of board seats isn't a crux.

A brief complaint about software: to uninstall my bugged GPU driver before a reinstall, I used AMD Cleanup Utility, which briefly booted my PC in Windows Safe Mode. MS Edge briefly launched in that mode for no reason. Then on restarting into normal Windows, I discovered that my MS Edge cookies and settings had now been nuked. Apparently that's a known issue. I had recently switched away from Firefox because Chromium is supposed to be a more mature browser engine, only to experience this joke. And of course the mode which wrecked everything is called "Safe Mode". The analogy to AI Safety is left as an exercise to the reader.