Curation is important. I force myself to do it by making it a manual process to move photos from my phone or camera onto the NAS. My phone and my SD card each has a small capacity so they'd easily be full if I don't curate.
When I'm traveling I try to do this curation the same day of taking the photo. In practice it's a great time at night to go through all photos of the day and discuss any interesting ones with your partner.
Also for this reason I almost never take videos. They are too troublesome to curate. Going through photos is quick. It takes only a second to decide to whether it is worth keeping. And once you decide it is, it takes less than a minute to touch up on the curves and the colors.
Because they're really different? system-ui is the specific font the OS uses for its interface... San Francisco in macOS/iOS/iPad OS, Roboto in Android, Segoe UI in Windows, etc. sans-serif is often Arial or Helvetica, but could be anything.
You seem to think GPUs are better than TPUs for rapid iteration. Why is that? There's no inherent reason why one is more suited to rapid iteration than another; it's entirely a matter of developer tooling and infrastructure. And Google famously has excellent tooling. And furthermore, the tooling Google exposes to the outside world is usually poorer than the tooling used internally by Googlers.
It was a few years ago since I last played with Googles hw, but iirc TPUs were inflexible and very fast. Worked well for linear and convolutional layers but could not accelerate certain LSTM configurations. For such networks GPUs were faster. It wouldn't surprise me the least if TPU hardware support lagged behind what the latest and greatest LLMs require for training.
It's "made by" TSMC as usual. Their customization comes from identifying which compute operations that want optimized in hardware and do it themselves. And then they buy non-compute IP like HBM from Broadcom. And Broadcom also does things like physical design.
My push-up story: I used to think I can easily do many pushups until I met my personal trainer and did push-ups in front of him. In his opinion, none of my pushups were correct. It was discouraging to hear someone dissect my mistakes: elbows flaring out instead of being held close to the body, hand position too wide or too forward, hips dropping. Of these, fixing the last mistake of hips dropping seemed the most elusive. Some days I could control the hips perfectly well, but other times I seemed to be unable to control the movement of the hips.
I still do 50 pushups per week with a mix of good and bad form (which is of course way less than 10k per year), but I found that I've mentally associated pushups with an exercise that I couldn't do well. It doesn't give me any dopamine boost. I'm much happier doing something else like barbell squats which I could do with good form and increasing weights.
Don’t be discouraged, the concept of a “correct push-up” is made up by humans. There’s nothing that makes that definition the truth, except that we’ve agreed upon it. To be fair, it probably works out your muscles more effectively, or in a balanced way or something. But even your “incorrect” push-ups are correct for _some appropriately defined cost function_. I promise, what you are doing is optimal. I just don’t know what for.
And doing 0 pushups is 100% worse than doing 50 of your personally-defined push-ups
Sure. As long as you ignore the push up forms that develop problems when done over long times. Not all "that's not the correct form" comes from clueless gatekeepers. Sometimes it comes from actual experts who know certain repetitive motions can lead to injury.
Ya, bad form _may_ cause problems. But I'll claim that many more problems arise from doing nothing and getting really out of shape, rather than from doing one of the "wrong" things.
Besides, I can kick my (rather low) ceiling, and none of your experts will advise me on the correct form for this. Without their advice what am I supposed to do, just stop kicking my ceiling? Ridiculous.
I don't necessarily disagree with your overall message. But I do think you're undercounting the number of folks who get injured through ignorance. And I don't even necessarily mean traumatic injury. Even just enough injury or soreness to discourage them from trying things again can be problematic enough. Is kicking your ceiling really a good measure of some sort of fitness? Is it worth the risk of slipping and cracking your head? I don't know. Maybe. But it's far more risk of injury than walking around the block a few times and the vast majority of people really just need to start there.
No, you can easily cause injuries from bad form that you wouldn't have gotten simply by doing nothing. I know several people know who broke or damaged something in their bodies and a few who needed surgery to fix it due to bad form over time.
I again claim: many more problems arise from doing nothing and getting really out of shape, rather than from doing one of the "wrong" things. You know several people that have been injured, i don’t doubt this. But in the US, 40% of adults are obese[1]. Europe is better but still bad (~20%).
Obesity is rather unrelated to exercise, it is about diet primarily; one can do nothing wrt exercise and still be thin. Injuries in terms of suffering are far worse than being unfit (not necessarily obese but not exercising in general), the amount of pain is incomparable.
While I agree that bad pushups are better than no pushups until you fix them, your first paragraph is misinformation. Anatomy is a 2,000 year old science, not a personal opinion. As a point of fact, if someone flares out their elbows when doing pushups, this will cause two things. #1, it will increase recruitment of the chest muscles at the expense of the triceps, which isn't necessarily a problem. But #2, it will place stress on the shoulder joints, and over time, this can lead to shoulder injuries.
There is no good reason to do pushups in a way that will cause joint problems. If you want to build your chest with pushups, the answer is wide grip pushups, not flared elbows. Your parent may find they are quite good at wide grip pushups, this is common because the chest muscles are bigger and stronger than the triceps.
Given sufficient time, a trainer worth their salt will teach you to develop every muscle. They might very well have you do wide grip. But there are definitely incorrect exercises, those are the ones that harm you. Dead lifting with rounded shoulders is a classic example, lots of people ruin their back doing this. I do both standard and wide grip push-ups. Though personally to build the chest I prefer the bench and other forms of chest press.
Ah... so I always though pushups were for my chest, but my triceps always got a workout. So I thought I was doing something wrong. I have shoulder issues so the tricep pushup is probably better for me.
I split my push-ups usually between 50 normal ones and 50 "wide" ones to get the pecs. Do I do them correctly? Hah. Probably not half of them. I try to keep myself planked, but of course my stomach slips toward the floor sometimes. Generally just make sure you're actually using those muscles. Forget about a coach. You know if you're doing it or "cheating".
Instead of 50/wk. Try 10/day. Just from my own experience, there's a benefit to a short concentrated burst of activity on a daily basis, versus a longer one on a weekly basis.
If you do hundreds of half assed pushups every day, consistently, you'll still be able to do 100 perfect pushups in one set when some pedantic "those aren't pushups" type tries to make you feel bad for finding a path to fitness without being a form nazi.
All that matters is area under the curve - do them daily and in quantity. Hungover? do them. Sleep deprived? do them. Headache? do them.
Form is so far in diminishing returns land it's hilarious that people bring it up, and in my experience often what's considered "perfect form" actually puts more strain on joints and increases likelihood of injury.
Wouldn’t the personal trainer be questioning his own relevance to your agreement if you were performing your push-ups correctly right from the start?
After all, as soon as he points out something you can improve on and tells you how to do it better, you receive positive feedback from him and the reassurance that you didn’t hire him for nothing.
And regarding push-ups themselves: isn’t it rather one-sided to train only those limited muscle groups with such a high number of repetitions, to the point that it leads to anatomical imbalances?
Don’t feel bad the other day a friend sent me a story about a guy who recently set some record by completing 1,721 pushups in one hour. Needless to say I dropped down and did 20 more.
A small book I read years ago mentioned said Bruce Lee would do 1k daily as an example of self-discipline. It was in addition to all his other training, this was just the baseline.
Once you can do 100 in a set though, that's just ten sets throughout the day. Totally doable if your employer can tolerate a sweaty employee.
> elbows flaring out instead of being held close to the body
not sure about others, but I think this one is already controversial, having elbows close to the body shifts load to triceps. Canonical form is to have arms 45 degrees to body in kinda arrow form, this will engage shoulders and chest more.
> It's worth being careful to only add indexes that will be used by real queries.
This reminds me of a technique used by Google App Engine SDK a long time ago before it was called cloud. Basically in development mode, the SDK captures the kind of queries you make, and then automatically add any index that would speed up this query into a configuration file. You then later deploy with this configuration file, which tells the production data store to create such indexes. I thought this was a genius idea. Not sure if something similar exists for SQLite.
It's not quite the same as capturing all of the queries used in development (or production), but it seems somewhat useful.
I'll also note that I had an LLM generate quite a useful script to identify unused indexes (it scanned the code base for SQL queries, ran `EXPLAIN QUERY PLAN` on each one to identify which indexes were being used, and cross-referenced that against the indexes in the database to find unused ones). It would probably be possible to do something similar (but definitely imperfect) where you find all of the queries, get the query plans, and use an LLM to make suggestions about what indexes would speed up those queries.
I also made several hobbyist apps for myself. Sometimes I thought what if I just posted it as a Show HN, but I am just rarely in the mood for implementing features others ask for, outside of work which pays me enough that I'd oblige.
It's been 25 years now but this exact thing is why I switched majors away from CS in college, I decided back then that I didn't want an employer/client telling me what to build. I liked software as a hobby but think it would have been a miserable profession (for me, not projecting to anyone else).
This is our code. That is their code. Depend only on the interface of their code and not the implementation. You can look at their code for curiosity but don't depend on the implementation of their code in our code.
Then you don't care what subset of the language their code is written in.
This fails to address the question you replied to. When "our code" is memory-safe and "their code" isn't, it's still "Our Product™" that ends up with a CVE that costs us money and reputation, because users don't care about whose code it was.
It also triggered some memories for me too. A college professor wanted to teach all the bit manipulating stuff and gave an assignment where students had to transform branchy code into branchless code using shifts and bit operators. Had a lot of fun doing that.
Yeah this is the kind of thinking opening up new heights when you get it!
These type of exercises should be mandatory for all compute intensive related jobs, especially for all data science people (i know some of them know this stuff but that does not seem to be the majority)
When I'm traveling I try to do this curation the same day of taking the photo. In practice it's a great time at night to go through all photos of the day and discuss any interesting ones with your partner.
Also for this reason I almost never take videos. They are too troublesome to curate. Going through photos is quick. It takes only a second to decide to whether it is worth keeping. And once you decide it is, it takes less than a minute to touch up on the curves and the colors.
reply