"for james watson, dna was everything. other than racism and sexism, which he was also a fan of, and which one might hypothesize could have lived upstream of the dna obsession."

i did some donation goal self-care things once and people got pretty into it. it's fun watching someone you like have a little relaxing time sometimes.

i think it's just sensible if you grew up with it cause you're used to it and there isn't much more to it than that, but stuff mapping well to base 10 is certainly not the usual criteria for measurements used in america.

at the end of the video i go over the full answer key for the test, btw.

SO MANY WORDS. i started self-imposing conditions that i'd only play colorless cards, or red cards, or w/e, so i didn't have to read as much.

the three components to making an llm work are
1. researchers who work out the tech for it
2. creatives who create the input for its learning models
3. investors who bankroll the tech

the thing we're getting most wrong is rewarding [3] and looking to them for expertise, while underutilizing [1]/[2]

or that there are fewer moving pieces for. like, it's much much much easier to land a spacecraft on the moon than it is to understand how ocean currents respond to changes in global temperature.

there's an awful lot of stuff we can either convincingly study ceteris paribus, that we see happen regularly, or that we have actual agency over as humans.

there is a way to drive through this crowd of people--but it is strewn with obstacles

if the pacific ocean breaks then sure, that kills humans. but also some wetlands can break and kill endangered species and that is actually a bad thing too.

oh and also a lot of the things that break are just completely disregarded. a lot of the "damage" numbers consider ~[human life and property damage and zero other things]. realistically, the ones we act on care about even less than that, and focus on first world damage and loss of life.

and how do you even publish a scientific paper which says an ocean current will stop working if the temperature increases 1 degree. like, where does the scientific method tell you how to do that. this isn't a thing we can study ceteris paribus, so now you have to get into statistical forecasting.

so the 3 degree at 2050 number ends up being ~"this is a good response to all of the near-certain science we have" but does not consider "this is a giant fucking planet and we don't fund climate science all that much and some of the shit we haven't studied will break at an undetermined time".

also our method of inquiry and conversation about this is resistant to people broadly understanding that. if someone does some research and finds with 90% confidence that a tidal current will stop working at 1 degree celcius increase, there's a LOT in the way of us updating to consider that.

we know that there's some point where we're very sure a lot of shit will go wrong. we do not have a thorough understanding of how much shit will go wrong at points prior to that. there are a lot of natural systems on our planet which could go very wrong, some of them very rapidly.

github.com/Card-Forge/f...

do you think maybe he is cheating such that he can tell when they're going to intentionally walk him

many will tell you about the card "claw". do not, under any circumstances, give in to the allure of this card. it is a deleterious undesirable and no amount of dealing 3 + 2n damage will heal your damaged soul.