5ellaal5

Currently I have three special interests and few filters:

queer polyamorous healing sex headcanon-happening in the Horizon game series and Hozier's music (not involving the artist himself) that I write way too many words about, and some people seem to like

Modern Monetary Theory and how it can empower us to defeat Moneyland and the horrifying corruption inflicted on us by Corporate Capitalist Empire (which needs to end) (see starter pack resource list below). If you’re not into that, you can block the tag “End Empire”

a lens of understanding interpersonal (and systemic) conflict through Empathy and Collaboration, or people (and systems) that are Empathy-Challenged and Non-Collaborative (this one is sort of my day job so I’m not revealing it here but I left pretty big clues I guess)

Two things I deeply appreciate:

2nd one first: Chris Hayes parrots a use-case for an enterprise (company’s subscription rather than individual subscription) AI agent to replace a 1st year law graduate at a law firm, presuming that all they do is compose and review documents.

Ed Zitron (guest) immediately counters that with “people don’t know what people do at jobs” and so can’t actually program a replacement, and the desire for Artificial General Intelligence is to have a slave.

1st one (chronologically in the episode): Ed explains the current economics are all an exaggeration of what’s been created and normalized over the past several decades of international banking and incorporation:

Making shell companies to sell things to which then sell them back to yourself in a merry-go-round, because each round is logged as Profit.

Invidia, the AI compute chip makers, have an actual product they sell. Then they take that money from selling a physical thing, and loan it to start-up data centers with very poor collateral or chance of future profit, so that the data center can… buy Invidia chips, which gives Invidia the “profit” to do it again with another shaky data center.

Everything else about the hyperscalers is big tech loaning each other money to rent Compute (use of a data center) and then loaning each other the loan or buying each other the compute, to prop everything like an Escher drawing.

Trillions of dollars of “investments” and also loans (debt) created, and if you remove all the people handing each other money and handing it back, there’s only about 1 billion in “profit” and that’s Invidia’s physical objects (chips)… which will have to be replaced because they are obsolete or burned out by the time a three-year construction project actually finishes.

Trillion: 1,000,000,000,000

Billion: 1,000,000,000

(The difference isn’t 1,000. It’s 1,000,000,000 a thousand times)

Last thing in general: I agree with Ed Zitron and also the DAIR Institute folks that AI boosters need to be banned from talking about what AI will, could, might be able to do — the future tense. Only listen to what it’s capable of now.

Below the cut is some of the transcript where Ed calls out that the goal is slavery. I would include more but I lost a bunch due to formatting glitches on my phone’s Tumblr app.