Jach's personal blog

(Largely containing a mind-dump to myselves: past, present, and future)
Current favorite quote: "Supposedly smart people are weirdly ignorant of Bayes' Rule." William B Vogt, 2010

Better world proposals as admissible heuristics

In CS, and graph searching in particular, the concept of an admissible heuristic is one that never overestimates the true cost of something. (Typically achieving some goal.) Different heuristics may give different estimates, but admissible ones never overestimate.

The goal of utopia is a perfect world, or at least a perfect-as-possible world. Dystopian fiction contains great fictional examples of how some would-be utopias aren't actually all that great. But I contend that a lot of those dystopias are still actually better than the present world, overall, and that reaching the perfect world may require such stepping stones. I worry that dystopias can represent local optima and thus be worse in the sense of cutting off the possibility for improvement, but I'm not sure that's possible on a global scale for all time.

Thus it's important to remember that proposals to make this world better, or ideas and visions of what possible future worlds might be like -- say an ill-defined World Without Suffering -- aren't proposing the ultimate perfect utopia, but merely improvements. And if they are better overall, and don't try to pretend to be perfect and final, then we can consider them admissible... To take the previous example, perhaps a certain amount of suffering is needed for human existence to have meaning. However the world is currently full of much suffering that I think we would be better off without, and once we are without, then perhaps we can reason on a further improvement to introduce the right amount back in, which would be another overall improvement on the path to perfection and thus admissible too.

See Full Post and Comments

Nim project

Finished up this fun little side project introducing myself to Nim (and SDL2 in the process):

See Full Post and Comments

In favor of privacy, but not as a right

I don't really believe in "rights". I believe in assurances granted by others, and when those others happen to be governments, whether it's a "right" or a "law" matters little to me. But other conceptions of "rights", don't buy it. If you try to argue some rights are objective, or even self-evident, I don't buy it even harder.

I still think many (though not all) the things I supposedly have rights to are nice to have, though, but not for the circular reason that rights are good.

When it comes to privacy, I generally fall into the "none of your/my damn business". There are many things I or you simply don't need to know, and I'll get ticked if you start trying to learn those things, and I'll understand if you get ticked in the other direction. For instance, say you're visiting my blog, and my blog asks for your browser to share your location (which may be from a phone, and thus very accurate). This is none of my damn business, I'm not trying to serve you software that makes use of mapping, something whose business legitimately is interested in your location.

See Full Post and Comments