> Do you have any proof whatsoever that non-programmers are using LLMs to write small bespoke apps for them successfully?
I’m radiologist, I’ve been paying for software that sped up my reporting like 200 usd per month. I’ve remade all the functionality I need in one evening with cursor and added some things that I’ve found missing from the original software.
> I haven't read the underlying paper so maybe they addressed this, but: couldn't this just be because sedentary/unfit people tend to drive everywhere?
It could be and that's the point - the exact causal effect of fitness might be overstated because some mortality reduction might have stem from - for example - driving less.
Yes, and since there is a factor that affects both fitness and dying in car accidents or homicides, then it's harder to establish for how big reduction of mortality is fitness really responsible (causal). What you are complaining about is exactly the point of this study:
> This suggests that people with high and low fitness levels may differ in other important ways, which is something that previous studies have not fully taken into account
It’s more like you are supposed to toss 1000 times and after 500 tosses you get a lucky streak of 5 heads in a row and then decide to end experiment and conclude that coin is biased.