Scrolling over HackerNews (HN) two weeks ago, I came across a post entitled "Two days of oatmeal reduce cholesterol level". The HN comments included stories from users about how their lives have changed for the better (e.g., fixed cholesterol problem, determined preference for soybeans), as a consequence of consuming oats.

The linked study Cholesterol-lowering effects of oats induced by microbially produced phenolic metabolites in metabolic syndrome: a randomized controlled trial, looked the effects on cholesterol, the gut microbiome, and other related factors of having non-diabetics with metabolic markers indicating increased risks for diabetes eat 300g oats + (optional) fruit/vegetables for several weeks. Finding from Abstract:

Both oat diets lead to an increase in plasma ferulic acid (0.64 [0.26, 1.02], P = 0.002; 0.55 [0.21, 0.89], P = 0.003), while the high-dose oat-diet also increased dihydroferulic acid (1.23 [0.44, 2.01], P = 0.003). Here we show that microbial phenolic metabolites are driving factors for the cholesterol-lowering effect of oats, which might be of relevance since short-term, high-dose oat-diet is a suitable approach to alleviate obesity-related lipid disorders.

The "headline" finding was that 2 days of exclusive oat consumption reduced LDL cholesterol by ~10%.

Upon seeing both the HN discussion and reading over the paper, I decided to (1) restart my oat consumption, which had ceased for several years, due to a fairly consistent feeling of lethargy and tiredness from eating the oats that would take several hours to overcome (I ate oats alone, with blueberries, or with honey) and (2) to see what I would happen if I replicated to some degree the conditions of the study, i.e. I only ate oats for an entire two days.

Regarding (2), I made it up until dinner on the second day before breaking from the 2-day oat diet. I found that (a) I developed a lot of gas, (b) I was surprised with how satiated I felt, (c) I felt that there was a lessened latency with my movements, in the sense of not being weighted down by digestion, (d) I consistently, over the 2 days, felt mildly tired or inhibited. I ate 3 cups of steel cut oats on day 1 and 2 cups on day 2.

Regarding (1), 2 weeks since making this decision, I have consumed 1 cup of oats on average every 2 days. There is still some gas, but not nearly as much as on the initial 2 days involving a total 5 cups of oats. I reserve the oat consumption for evenings, since my work is less compromised then by the feelings of lethargy and lowered motivation engendered on occasion by the oats.

Using Perplexity to examine why I might feel lethargy as a result of eating oats:

(1) "Tryptophan: One cup of uncooked oats provides 130% of the recommended daily intake of tryptophan, which converts to serotonin (the "feel-good" hormone) and then to melatonin (the sleep hormone)." (2) "Butyric acid precursors: Whole grains like oats promote production of butyric acid, which helps the body synthesize GABA—a neurotransmitter that calms the central nervous system"

Some evidence from research in support of the claim made in sentence 1:

We aimed to compare the effect of a 10-min walk immediately after glucose ingestion (10-min walk condition) on glycemic control to that of a 30-min walk, 30 min postingestion (30-min walk condition). In a randomized, crossover, counterbalanced trial with three (control, 10-min walk, 30-min walk) conditions, twelve healthy young adults (6 females) walked at a comfortable speed during the walking conditions (control condition = rest) after glucose ingestion (75 g). The walking conditions yielded significantly lower 2-hour glucose areas under the curve (10-min walk = 15607 ± 702, 30- min walk = 15732 ± 731, control = 16605 ± 745 mg·min/dL) and mean blood glucose levels (10-min walk = 127.9 ± 19.4, 30-min walk = 128.9 ± 5, control = 135.8 ± 20.5 mg/dL) than did the control condition (p < 0.05, d = 0.712-0.898). The 10-min walk condition (164.3 ± 8.9 mg/dL) resulted in a significantly lower peak glucose level than the control condition did (181.9 ± 8.4 mg/dL, p = 0.028, d = 0.731) despite no significant difference between the 30-min walk (175.8 ± 9.6 mg/dL) and control (p = 0.184, d = 0.410) conditions. A brief 10-min walk immediately after a meal appears to be an effective and feasible approach for the management of hyperglycemia.

(from https://www.nature.com/articles/s41598-025-07312-y; emphasis added)

I am a someone who has participated informally in self-experimentation and in quantified-self activities.

Reading work composed by Gwern (e.g. this piece) and niplav's (e.g. this piece), alongside writing in the fields of medicine (e.g. this paper) and psychology (e.g. this paper), has been both illuminating and motivating.

I am trying to formalize my data-recording for SE/QS and have found that the following approaches have not "stuck":

• (1) automated emails or Google Forms
• (2) cron-tabs on my MacBook for digital EOD entry of data
• (3) alarms at EOD for manual entry of data into .csv, .txt, or .json files

One approach that has given me optimism and feels relatively more entertaining (i.e. that seems to "stick" better) has been the following: once per week or month I generate a $LaTeX$ PDF document (using Python to generate the document w/ configuration and command-line arguments) with checkboxes next to categorical variables I want to track (e.g. whether I have consumed at least 2 liters of water; these are loaded from a config.json file) and EOD I fill in the boxes with a pencil for what I did. The goal is to eventually be able to scan the PDF sheets and automatically extract the data for analysis.

Here is a section of an example sheet: here. Here are some of the variables: □ Reading for at least an hour and a half. □ Studying Anki cards at least once. □ Bathing at least once. □ Making at least one comment on an on-line forum. □ Consuming at least a coffee’s worth of caffeine. □ Going to bed before 1:00 AM EST. □ Drinking at least 2 liters of water. □ Meditating for at least 15 minute

I'm curious whether it's possible to turn a resource, like a math textbook, into flashcards in such a way that you memorize the flashcards themselves, enabling a purely mental flashcard review.

Somewhat related to this: to have Anki flashcards or some higher-level flashcard representation (which could yield an Anki deck, e.g. using a genanki workflow) , more consistently extracted from textbooks and made available would make me very happy; I've wished for a reality in which, on data or code sharing platforms (e.g. Zenodo or GitHub), the majority of textbook authors or researchers shared their .tex files (for others to use, possibly with Anki) or any Anki cards associated with their work.

Thank you for sharing some thoughts. Somes comments:

• [Question] Do you ever find yourself slipping into a state (possible gradually) of over-reliance on the LLM? This could instantiate itself as a momentary mental-rupture of the routine, where you find yourself pausing and realizing that now it feels more difficult to do something yourself that was, at a time in the recent past, moderately ordinary?
• [Question] As Claude and other LLM become (expectedly) more capable, do you have any preliminary thoughts for how this workflow style may be updated?
• [Remark] I find myself most entertained and invigorated by having Claude generate lists of posts on a particular subject that could be written, along with short descriptions of what these posts might contain. I've used the LW API to get LessWrong posts before asking of Claude something akin to "What is missing from this selection?".

I done similarly with my CLAUDE.md, including a line close to what you've written but also the line "Please search for and consider other solutions beyond those I've suggested before writing any code." It seems likely to me that there are better ways to communicate this notion to Claude, though, and I've done less experimentation than I would like.

Upvoted on the basis of clarity, useful / mentoring tone, and the value of the suggestions. Thank you for coming back to this.

In a first-pass read, there is not much I would add, save for mentioning that I’d expect (1)-(4) to change from what they are now were they to actually be implemented in some capacity, given the complexities (jurisdictional resources, public desire, participation, etc…).

I have the Myth of The Rational Voter on my shelf unread!

If I have any sufficiently useful or interesting ideas or comments regarding your remarks, I will add them here.

Agree. This post captures the fact that, time and again, historical and once perceived as insurmountable benchmarks in AI have been surpassed. Those not fully cognizant of the situation have been iteratively surprised. People, for reasons I cannot fully work out, will continue to engage in motivated reasoning against current and near-term-future-expected AI capabilities and or economical value, with some part of the evidence-downplaying consisting of shifting AGI-definitional or capability-threshold-to-impress goalposts (see moving goalposts). On a related note, your post also makes me imagine the apologue of the boiling frog of late w.r.t. scaling curves.