• 0 Posts
  • 143 Comments
Joined 1 year ago
cake
Cake day: July 14th, 2023

help-circle
  • Most recent social security trustees report says the trust fund will run out in 2035. What happens in 2035? Benefits are still funded at 83% in perpetuity. By the way, last year it was going to run out in 2033, and the year before that it was going to run out in 2031. And also by the way, the trust fund was specifically set up because they knew the baby boomers were going to stress the system, so it’s supposed to get depleted as the boomers use it.

    Everything is working mostly as intended, and yet there’s all this anxiety around Social Security. Why? Because Republicans want you to think Social Security is fucked all on its own so that you don’t question it when they ratfuck it. That and they want to constantly frame the conversation as such so that the conversation doesn’t turn to “how do we make social security more robust and generous?” or some other radical socialist nonsense.







  • Let’s just take NYT for example. Subscription costs $325/year. Why would I ever pay that much? It’s not 1954. I’m not sitting down with my morning coffee and reading the damn thing front to back. I’m reading maybe one article a week from 15 different sources. Am I supposed to pay $5000/year just to cover my bases?

    As with everything else in [CURRENT YEAR] the value proposition is so absurdly out of step with reality that fixing it basically relies on rolling out the guillotines.










  • I guess I’m wondering if there’s some way to bake the contextual understanding into the model instead of keeping it all in vram. Like if you’re talking to a person and you refer to something that happened a year ago, you might have to provide a little context and it might take them a minute, but eventually, they’ll usually remember. Same with AI, you could say, “hey remember when we talked about [x]?” and then it would recontextualize by bringing that conversation back into vram.

    Seems like more or less what people do with Stable Diffusion by training custom models, or LORAs, or embeddings. It would just be interesting if it was a more automatic process as part of interacting with the AI - the model is always being updated with information about your preferences instead of having to be told explicitly.

    But mostly it was just a joke.