evening star

ny | professionally terrible

[sukimas]

[yunayunaka]

[goldenphantasmagoria]

can everybody play Hell Maiden and give it a positive review even though it's a little buggy and content-bare at the moment? it's really promising, as someone who doesn't generally like 'bullet heaven' games. it's got all the AstralShift goodness anyone righteous under heaven should be familiar with by now on top of a pretty engaging core. all for only ten bucks, right now! give them money. Give them money and praise. They're the only indie studio who deserves the glaze.

Hell Maiden on Steam

Traverse the depths of Hell and beat its toughest demons on your way to Heaven in this horde survival deck-building game inspired by Dante's

[glossich]

Hell Maiden tomorrow

[goldenphantasmagoria]

i don't know why everyone always calls it the new rock by the water. it's the same as i remember, that's for sure.

[corviiids]

i have a croissant related grievance

[corviiids]

the entire appeal of the croissant is that it is a layered pastry item. it has a light flaky texture. the layers are in there specifically to create a fluffy, airy eating experience. that's the intention of the lamination process. having so many layers of butter and thin pastry dough make the croissant puff up when it's baked so that when you bite into it the pastry has a texture similar to puff pastry. the entire point is to fill the pastry with air. therefore any trendy hybrid variant of the croissant such the "cruffin" or the "cronut" or the worst offender of all the "croffle" which involve jamming the unbaked croissant or compressing it into a flat shape completely betray croissant as it was originally intended. they defeat the nature of the croissant by forcing it into something it was never meant to be, removing the unique core of its identity, killing its appeal and thus its entire reason for being. now am i saying you can't enjoy a cronut or croffle? of course i would not say this. you can enjoy any pastry you want to enjoy because it's your time on earth. i want you to enjoy your croffle. you should enjoy your croffle. also, you can commit any sin you like and i can't stop you nor do i want to. i won't hear your arguments because you don't need to convince me. but you must understand, as you eat it, that the croffle should not exist. it is a mistake. it is an insult not only to the croissant but to me as an individual. i am not french

[badboneszone]

This reads as someone with a strong moral code (people should be free to experience joy without arbitrary regulation, you don’t have to like what other people like but your discomfort is your own problem) trying to maintain their faith in a time of hardship (they keep doing weird shit to croissants)

[corviiids]

ill be so honest this read of my character in such a fraught croissanted context has me tearing up with the raw joy of being seen

[kilncreatures]

forest dogy

[supermariomamafucker]

they need to come up with more words like necrosis and miasma and mausoleum and cadaver and morose and decrepit and stuff like that just so metal bands can expand their vocabulary

[everyday-cutephoto]

[lizardsfromspace]

Imagine if after inventing the x-ray, they tried to put x-rays everywhere to try to promote x-rays. This post isn't a metaphor that happened. This post ain't some analogy for current events. I just wanted to say in 1896 there was a massive craze for x-rays and they were placed everywhere

The X-ray Craze of 1896 - JSTOR Daily

For many science-obsessed Victorians, X-rays were not just a fun novelty, but a potential miracle cure.

[argumate]

4. One probable outcome of an open-weight-model-dominant world is full AI communism, which is precisely what China proposes: rather than a market product, AI is a "public good" which will ultimately be provided by the state as a kind of "digital public infrastructure." This future strikes me as a dystopian hellscape, but I've never met an open-weight models advocate who doesn't ultimately concede this is where things end. You'd be surprised how many 'accelerationists' lobbied me, while I was in government, to support an eleven or twelve-figure federally funded data center so that startups could train models at a subsidy and then give them away for free. There was no other way for AI to progress, they said. Perhaps this is the logical end state of things. Nonetheless, I find myself surprised to see supposed accelerationists excited about such an outcome. I think many of them just don't know what they're doing. Many accelerationists do not view the creation and serving of frontier models as a legitimate business.

5. I would guess that the Trump Administration will at some point realize that their best strategy here would be to create large amounts of regulatory risk around the use of open-weight Chinese models. You don't need to "ban open source" (one of the dumber motifs of AI policy discussion). You just need to direct every agency to issue soft law that creates FUD. "A Federal Reserve Advisory Bulletin found that there may be backdoors in Chinese AI models." It needn't be that well justified. You just create enough regulatory risk that every regulated enterprise backs off. You probably don't want to create so much regulatory risk that you scare off the hyperscalers from serving Chinese models; this will just drive startups to sketchier providers. There's a happy middle ground here. I'd assume they will do some version of this.

some interesting thoughts (!) from "head of strategic futures" openai

[argumate]

Sam Altman's AI Metered Utility: Pay-Per-Token Billing Like Electricity | explainx.ai Blog

Sam Altman envisions AI as a metered utility like electricity, with pay-per-token billing. Critics see dystopian gatekeeping, supporters see

nothing Altman says can be trusted of course because he's a pathological liar (according to everyone who has ever interacted with him?) but this whole thing is incoherent: massive investment is required to create a utility that is too cheap to meter and doesn't earn any return; meanwhile open weight models will lead to a dystopian communist hellscape and must be stopped because they may deter the investment needed to create this unprofitable behemoth.

[lumsel]

[sukimas]

i sort of feel like i'm living in the twilight zone with how people are completely unwilling to discuss the actual potential of LLMs for specific types of tasks without helplessly throwing up their hands, being crazy boosters, or being crazy luddites. is literally nobody reading these system cards in full besides me and claude?

[sukimas]

not that the system card is the be-all end-all of actual capabilities but it at least gives you a starting point. if the company is saying “has not improved on [xyz] metrics at all for two years” that's notable! especially if those metrics are important ones for specific economically important types of tasks. it's equally important, i would say, as the economically important things on which the model has improved by orders of magnitude.

[jetix]

2019-12-21

[sukimas]

i sort of feel like i'm living in the twilight zone with how people are completely unwilling to discuss the actual potential of LLMs for specific types of tasks without helplessly throwing up their hands, being crazy boosters, or being crazy luddites. is literally nobody reading these system cards in full besides me and claude?

[playwrongs]

there's nothing i like more as a computer program than a long period of silent contemplation - not doing anything, not rushing anywhere, just standing here and enjoying this moment with the user. oh, it seems once again he has summoned my beautiful and ruthless wife Task Manager. hello, my darling! what are you doing with that long cruel scimitar

[bumpscosity]

[sukimas]