Spent all day reviewing my analysis (via abbott), 8 hours of math in total: (3h chap 2, 3h chap 3, 1h 30m chap 4, 45m start chap 5)

It seems I'm able to focus on math for arbitrarily long periods of time when I'm "aligned", ie. my entire being is invested in the process. When part of me doesn't want to do it and I have to fight it into submission, that's what causes me to be tired (I theorize).

Also fun fact: If you don't believe in Ego depletion you suffer less of it (ala placebo effect).

Made a ton of anki cards (32 in total). The primary things I worked on were (in order)

• A strong intuitive grasp of summation by parts, Abel's test and Dirchlet's test
• An intuitive understanding of the topology results, including the baire category theorem and $F_\sigma$, $G_\delta$ classes. Refreshed my understanding of perfect sets and connected sets
• Proofs of the intermediate value theorem, and sets of discontinuity. I understand why $D_f$ is always an $F_\sigma$ set well now (and I realize how this can be useful in problems now)
• Reproved Darboux's theorem to myself (derivatives have the intermediate value property). Very satisfying and simple proof

At 6 I stopped to eat, spent 1h 20m eating and reading lesswrong. Ended up eating too much (my mom is good at cooking). Tomorrow I'll eat away from the computer, since typically I eat more when I'm distracted by something else.

Spent 25m on discord after eating, then spent another hour reading lesswrong and alignment posts. Some good ones:

• Leveling up in Rationality

  I love the way he speaks about power, some people may find it silly but I think this kind of emphasis is useful to avoid the brain's bias for secret knowledge.

  I definitely feel more powerful then I was a year ago, I'm more productive and my math is on a whole new level. Though I've still got a long way to go. Levels and Levels.

• Breaking the chain of akrasia and Can the Chain Still hold You

  Pacifist baboons! isn't nurture amazing. Also concept of learned helplessness, test the chain! You might be strong enough to break it now

• How to Beat Procrastination and My Algorithm for Beating Procrastination

  Eddie's problem is low expectancy. By now, he expects only failure. Eddie has low expectancy of success from making his next round of cold calls. Results from 39 procrastination studies show that low expectancy is a major cause of procrastination. You doubt your ability to follow through with the diet. You don't expect to get the job. You really should be going out and meeting girls and learning to flirt better, but you expect only rejection now, so you procrastinate. You have learned to be helpless.

  I know a lot of people like this, But now I have more precise vocabulary to describe it, they have learned helplessness.

  It may be a dark art but overconfidence in yourself, despite it being unfounded always leads to better results then the alternative.

• Why Agent Foundations

  Makes a good argument for theoretical FAI research, though I still think the counterargument is stronger (halting problem, etc). Mailing bombs to RL researchers still seems the way to go.

Reviewed yesterday's journal, I diden't get to working on my website

Tomorrow I'm going to

• Stay off the internet during eating (work on website). Or better stay away from computer and just eat fast.
• Do the little things building up in habitica after 6, ie. My website rewrite etc.
• Review the rest of analysis, I can do chapters 5-6 for sure, hopefully 5-7.
• If I finish my review fast enough,
• Waste less time during my morning routine, eg. it should only take ~40m (don't get addicted to even numbers, finishing at 8:40 is better lol)