There are definitely long books I've read where I didn't feel like I got enough out of them for it to have been worth it, like I read (listened to) Gravity's Rainbow right after and a lot of the characters and plot lines just annoyed or disgusted me, but I'm sure that book means a lot to a lot of people. In the interview "Of course you end up becoming yourself," the author talks a lot about how in avant-garde literature you have to make the payoff good enough to reward the reader for doing the work, and that equation is going to be different for everybody.
I just implemented this with oauth - basically I have a tool that returns a signed URL that the client can PUT the file to. Works in ChatGPT Work mode. My use-case was to have the client generate an image and upload it.
The end notes are interspersed throughout the book where they're referenced, and then a bell rings before returning to the main text. That's why I listened to it twice. I remember towards the end of the first listen-through, I sort-of recalled some of the earlier end-notes, like for example Himself's filmography, which I had basically ignored the first time through, but listened to intently the second time through.
Just finished a David Foster Wallace kick. I listened to the new 30th anniversary audiobook of Infinite Jest (twice) and then The Pale King (twice). Then I listened to "Of course you end up becoming yourself" which is a long-form interview with Wallace by David Lipsky, traveling with him on the Infinite Jest book tour.
I had to listen to the books twice because there are a lot of details that you won't get the first time through unless you're extremely intentional and detail oriented, but I found it was worth it to listen again. I'll probably listen to Infinite Jest again sometime soon, even though it's 64 hours long. IJ is actually really great sci-fi with some important questions for our time.
I'm not trying to say that other apps don't also provide great advice (especially RP, I love watching Dr Mike and have learned a ton from him), but I think that a lot of the flexibility in the programs in these apps are primarily for marketing purposes (like are you really gonna look like Black Adam after following the Black Adam program? unlikely..)
My marketing is for a different niche - people who lift weights, see that they have variability set to set and day to day, and want a program that will help them quantify that variability and understand what it means. One thing that works really well in my program is detecting when you need a rest or deload, because the baseline comparison is rigorous and sensitive - if you have one bogey set, it might not mean much, but when you see that even your opening sets are at or below baseline after a stretch of workouts, that's a pretty good signal, and that's hard to get with most apps.
I don't intend to monetize, I only want to compete for data. I think it's an interesting field and I'd like to be able to contribute back to the science by using the data gathered. I think these workout apps try to be opinionated about what an effective program is, but they try harder to make money, so they generally bend over backwards to accommodate whatever program the paying customer wants to do. Mine is more opinionated. I think this is the best way to lift weights for a wide range of athletes who are interested in getting stronger and bigger with the least amount of wasted effort and lowest chance of injury, the simplest way to record it, and the best way to get a statistically defensible understanding of how your strength is changing over time. If you want to do something different, as you mention, the space is crowded with apps.
physiological papers. My main interest is contributing to the research by developing a fatigue model. My hypothesis is that athlete recovery factors like accrued fatigue in-day and cross-day can be used to predict workout log outputs of total tonnage, combination of weight and rep count, and proximity to failure. Eventually I hope to be able to link these recovery factors directly to provide targeted advice about your overall program's effectiveness, deload planning, etc. in order to most efficiently achieve strength, endurance, and hypertrophy goals.
yeah, point taken, I do need to rewrite the user-facing documentation to be more accessible (and less obviously LLM-assisted, heh - although I did review and edit all of it). Workout design (including teaching a user how to do an exercise) and logging are two legitimately different tasks in a workout app and I wanted to focus on being the best logger. Maybe this is paranoid but I felt that saying "here's how you do a workout" introduced a little more potential liability than punting to "if you don't know how to setup a workout and learn to do the movements, hire a personal trainer to help you figure out movements appropriate for your fitness level."
This is really awesome! I used to have an movement to tissue mapping in an earlier version of app but I opted to focus more on the weight-reps curve it because the tissue volume calculations added too much complexity to the interface, and it just encouraged me to do too many different movements. So exercise selection and tissue coverage is something I actually think about outside my app.
I had coefficients mapping each exercise to its tissue impact but it was kind of arbitrary and unscientific - even between athletes you will have different techniques on the same movement that will impact tissues differently. I do think that workout design is a super important area and I hope you crack it, it was too hard for me to do a good job on.
yeah, it's a difficult balance to explain why this app doesn't work like every other weightlifting app and not inundate people with too much information. It actually is pretty simple once you start using it. I'll look at the tutorial scrolling issue, and eventually rewrite the content.
The tutorial can't be escaped because it requires a liability waiver checkbox at the end.
I love these! I recently designed a few decks of playing cards of mathematicians for my dad as a gift, my favorite part was designing the backs: a zeta function for the pure-maths deck, a Moore curve for the computation one, and two interfering waves for the physics one.
it's a weight-training app that helps you train along your "pareto frontier" of weight vs reps. The idea is to train at lower weight-higher rep, medium weight medium reps, and higher weight, lower reps for every movement. I tried to develop my own weight training program following bits and pieces of advice from bodybuilding forums and ended up injuring several tendons in my first year. So I did a bunch of research on tendon strengthening as well as what's most effective for hypertrophy (reps near failure) strength (reps near maximal load) and injury-prevention/frequency (not bringing yourself to failure too often) and designed an app to automatically prescribe and advance weights and reps based on your learned strength curve (Brzycki-like, with an added shape parameter)
The app is designed to make use of the free Cloudflare tier, so I can support thousands of athletes for just the cost of the domain name. I'm primarily interested in understanding the "Fatigue curve" - right now I have some basic per-set fatigue modeling (basically a log-linear strength dropoff) but I think it could be much better characterized with more data. I could go on and on about the modeling but my intention is to keep it free (maybe add some non-intrusive ads on content pages if it ever starts costing me a few pennies a month) but my primary interest is to be able to do statistical analysis on the data.
yeah, I have been amazed at what I could do for free, and then amazed at how much more powerful it got for 5 bucks a months. Cloudflare is killing it in terms of value for small websites (and features and reliability).
I haven't tried asking it in Hangul but these particular artists (and the photos I'm using actually) are linked to their romanized english names on e.g. Fandom so it's not unfindable on the internet
3.7 flash was by far the best model for image recognition tasks according to my benchmarks. 3.8 flash didn't regress any candidates and improved some specificity (positive ID of common name vs species name of exotic fruit, correct identification of cast/replica of artifact and statue) but is still relatively weaker (26/30) on esoteric public figures (Korean beatboxers). I'm going to have to make my benchmark harder.
My benchmark is tiny compared to WorldVQA or FG-BMK, which are available, so I'd point you in that direction if you're interested in a VLM benchmark. My use-case isn't exactly captioning as in "what is in this image?" -> caption, I am using the VLM to validate captions, as in "is this an image of [supposed subject]?" - my ranking of models I've benchmarked is gemini-3.7-flash > seed-2.1-turbo > gpt-5.6-luna > qwen-3.7-plus > qwen-3.7-flash. Gemini is almost perfect on my test dataset, only failing on some esoteric pop-culture minor celebrities and being over-specific in some cases (i.e. Q: is this [common name of fruit]? A: that's a [latin species name of fruit], not a [common name of fruit]; false). However, gemini-3.7-flash is only in my test list because openrouter has it on 75% introductory discount; otherwise it would be about 4x more expensive than seed.
I just ran my image recognition benchmark on it ("is this XXX public landmark"?) and it misses a lot that bytedance seed 2.1 turbo gets right; for example:
Asked "Is this Salisbury Cathedral" and supplied a picture of Wells Cathedral, it answers "Yes, the west facade of Salisbury Cathedral". Bytedance seed 2.1 turbo correctly says no. Similar results for a picture of Manhattan Bridge sent as Brooklyn Bridge, Chartres Cathedral sent as Notre Dame, etc.
I have a benchmark of 12 such images and seed gets 11/12 and deepseek only gets 6/12.
the only thing that still kind of annoys me is constantly being told what something is not, but even that statement is load-bearing (see what I did there) because it records how it ended up with this decision, because it's not that other choice that it mentions.
FWIW, I also think the constant chorus about how new models are worse than old models is a human hallucination. They're certainly not perfect but every one becomes more steerable in terms of actually completing more and more complex work.
I don't know that much about math but I've read that one possibility is he had found a valid proof for n=4 and assumed that it generalized. Hopefully someone else who knows more about the subject chimes in!
There’s no particular reason it’s not OSS, but my main interest is collecting a lot of data on different athletes and publishing original research. Most weightlifting studies are small n and over a short amount of time. My particular interest is how volume, load, and fatigue are related to strength, endurance, and compliance over time. My intention is to run it for a while, look at the data to generate some hypotheses, pre-register them, then run some experiments (and by that I mean just keep collecting data). If someone else was particularly interested in this goal, I would definitely invite them to the project. That’s why it was important for me to design it to be hosted for just the cost of the domain name, because I don’t really intend to make money from it, I’m just interested in the data.
Some nuance here: the latest research shows that proximity to failure is the main hypertrophy driver regardless of load and rep count; high rep count makes proximity to failure harder to gauge; so high load/low reps close to failure is probably better for hypertrophy (there are other good reasons to do higher reps/lower load work though)
I wrote this app as a SPA! It uses a curve formulation similar to Brzycki, except I added a “shape” parameter (an exponent gamma between 0 and 1) that slopes the 1rm downwards at the right side.
My main finding for “pick whatever weight you want today” was that picking a lot of different weights made the curve less identifiable, so my latest iteration encourages you to pick a ladder for a few sentinel exercises per mesocycle in order to improve the statistical power. In addition, strength improves more quickly at >80% of 1RM, and hypertrophy depends on proximity to failure, so if you pick a lower weight, you really need to go to failure, which burns you out for the rest of your session, where leaving 1-2 reps in reserve is probably sufficient for hypertrophy and leaves a lot more gas in the tank for the rest of the session. Definitely open to suggestion/discussion here.
https://curvefit.app (it runs on Cloudflare free tier, so I won’t have to start running ads or charging until I hit a couple thousand users)