The interesting question is how to define 'average'. Over what probability distribution?

Hey, anyone remember this from earlier in the week? https://daringfireball.net/2026/08/anthropics_watermark_text...

https://en.wikipedia.org/wiki/IBM_and_the_Holocaust

I fully agree with your base perspective, that the main focus should be on preventing authoritarians from coming to power in the first place.

I also agree with your final point, that the government doesn't need tech to lean on a person.

We disagree in the middle, on how tech that makes evil orders of magnitude easier might be a contributing factor and might enable the government to lean on many more people than it would otherwise.

I want to agree with you, but wonder if you have ever been poor? When you need the car to get to work so you can feed your kids, but you can't afford all of

- feed kids - rent - insurance

because you got hit by a surprise medical bill (kid got sick, maybe?)

I'm strongly in favor of your end goal (less car-dependent life), I'm just cautious about using punishment as a way to get there.

Unless we made the fine proportional to income?

Building the family's house alongside Dad sounds much more valuable than anything they would have learned playing on the school's soccer team.

The idea starts to make a lot of sense once you look at its framing. This includes the argument you make above (why can't you hire help out of pre-tax income, like any other business can?) and more on the legitimacy of taxation.

https://lawliberty.org/book-review/georgism-revisited/

price point is a little higher than the cybertruck, but hey, probably worth it.

15

45-50

150

450/500

1500

5000

So it depends how you define "diversity". I can see how from your view (range of life choices) humans today are more diverse than most species. However, if you go back not so long ago, there are species of ants with more worker roles (40+)than your typical midieval village.

https://arxiv.org/abs/2303.14151

Double descent is a surprising phenomenon in machine learning, in which as the number of model parameters grows relative to the number of data, test error drops as models grow ever larger into the highly overparameterized (data undersampled) regime. This drop in test error flies against classical learning theory on overfitting and has arguably underpinned the success of large models in machine learning. This non-monotonic behavior of test loss depends on the number of data, the dimensionality of the data and the number of model parameters. Here, we briefly describe double descent, then provide an explanation of why double descent occurs in an informal and approachable manner, requiring only familiarity with linear algebra and introductory probability. We provide visual intuition using polynomial regression, then mathematically analyze double descent with ordinary linear regression and identify three interpretable factors that, when simultaneously all present, together create double descent. We demonstrate that double descent occurs on real data when using ordinary linear regression, then demonstrate that double descent does not occur when any of the three factors are ablated. We use this understanding to shed light on recent observations in nonlinear models concerning superposition and double descent. Code is publicly available

If it is serious, you could always set a timer.

maybe not that one :)

some argue (using data and experimentation) that trees and/or other plants express all of these behaviors

The "upload csv file" box does not show up in the test project.

Our society thinks it is normal to have a creative work evaluated by accounting, but never the reverse.

It would be fun to reverse that. "Ok, Jones, you made a profit of 30M, but we don't really see that you released that much creativity. We're going to have to cancel your project."

yes, maybe because the regular store can buy in much higher volume, and will have faster turnover of stock?