Notices by James Tindall (atomless@mastodon.social)

@bruces Like all content creation with genAI, Vibe Coding is entirely different to the previous “democratisations” cited here. And this is what happens when capitalism disappears from the view of those operating within it. The piece argues Vibe Coding is merely the next phase in layers of abstraction that have grown over the years, between the software engineer and the bare metal of the machine. But while those were small evolutionary steps in capitalist alienation, genAI is a giant leap.

The utter debacle of the WWF tag team rant at Zelenskyy yesterday in the White House was a Shakespearian miniature that exquisitely lays out the culture war waged by the right against truth, legitimacy, authenticity, honour, and the real charisma and masculinity they so desperately yearn for as without which they have forever felt invalidated. 2/

That moment you realise your worst fears are confirmed, that this was never going to be about what you had hoped for, and was always already about framing you as a political patsy. 1/

Zelenskyy, is a true hero of the 21st-century, brimming with effortless authenticity, charisma, honour and strength. The cosplay duo wagging the finger and yelling at him from the right, embody only entitlement, insecurity and a desperate need to assert their daddy bought legitimacy by undermining all that embodies the legitimacy they lack. 3/end

I've been hesitant to publish this post for months. I feared my emotions would overwhelm any analysis I might muster. I hope I have managed, at least partially, to avoid that.

Engine Lines: Killing by Numbers
- the self-selecting users of the self-reinforcing prediction machines.

https://open.substack.com/pub/atomless/p/engine-lines-killing-by-numbers#

If you've not yet read @danmcquillan 's take down of Labour's AI "plan," and how it is a "gift to the far right," it is well worth your time.

https://www.computerweekly.com/opinion/Labours-AI-Action-Plan-a-gift-to-the-far-right

OpenAI: “this is a race America can and must win.”

AI always was capitalism’s new War Basilisk.

By that I mean not only does it operate according to the logics of war, but as a Thought Basilisk that invokes the future threat of being dominated by an imaginary ASI equipped hostile state unless we pour every resource at our disposal into supporting the few companies building next-token prediction machines.

https://openai.com/global-affairs/openais-economic-blueprint/

@mike805 @pluralistic I really like the detail of this story that it was Marx’s wife, Jenny von Westphalen, who pawned Marx’s trousers —apparently to buy food. Was there also perhaps an element of punishment to this choice? Were there really no other things to choose to pawn? Or was she actually helping him focus on finishing Kapital by removing the option of leaving the house?

How much we all benefit from @pluralistic ’s compulsion to write. So very grateful. Chapeau! https://pluralistic.net/2024/12/21/blockheads-r-us/

@kityates interestingly "Herrenvolk democracy" translates to "Gentlemen People Democracy". Maybe what we have now is more "Rich People Democracy" which in German would be "Reichevolk Democracy" ?

https://en.wikipedia.org/wiki/Herrenvolk_democracy

Ah, confusing.

@pluralistic Oh, I thought there may have been something that pre-dated Davies’s book? Especially as Chiang was already circling this early 2023 in his “Will A.I. Become the New McKinsey” piece in the NewYorker?

@pluralistic for citation purposes, which of your posts is best to link to for “accountability sink”?

@pluralistic @a_cubed @Axomamma @MattMerk

Case in point, in fact hard to imagine a more perfect example of a bubble of delusion arising through extreme concentration of capital that whilst in denial of its internal failures then operates to present a mirage of success.

@pluralistic @a_cubed @Axomamma @MattMerk

Enshitification, the Reverse Centaur, Accountability Sinks (or perhaps Impunity Ware), these concepts are extremely helpful. Feels like a quite coherent picture of 21st-century capital and its machinations is coming into view.

Just noticed the #treasure-what-you-measure anchor in the url which lead me to this old post I'd not seen before which is also excellent 😙🤌 : https://boingboing.net/2019/06/26/breaking-goodharts-law.html

@pluralistic @a_cubed @Axomamma @MattMerk
Thinking some more about the bubbles of delusion that envelope concentrations of capital and the attraction force they exert upon those without (capital). A significant factor here, especially in an attention economy and society of control whose labidinal flows are directed by the Advertising Industrial Complex, is concentrations of capital also have the power to distort the infosphere by overwhelming sense (of their failures) with a mirage of success.

@ariane @pluralistic
The infinite regress of management process, systems of labour control within a panoptic society of control, chimes rather well @baldur 's take on code frameworks and Generative AI being fundamentally about reduction of workers to the units of labour they perform, to maximise the power of arbitrage, while laying the way for full automation. https://www.baldurbjarnason.com/2024/react-electron-llms-labour-arbitrage/

This post from @pluralistic
is honestly up there with his concept of enshitification in the degree to which it explains so much of the shitshow.

It implies a compelling argument against great inequity. Namely that great wealth creates a bubble of affordance that insulates the rich from discerning the immediate consequences of incorrect or damaging notions or actions so delaying recognition of errors and the need to alter course.

https://mamot.fr/@pluralistic/113548449727704405

@danmcquillan @anildash I have misgivings with this critique.

To me LLM's unreasonableness comes not from any irreproducibility but from their reproducibility. The way they will repeat ad infinitum the statistically prevalent attitudes and biases locked into them via the spread of opinion within their pre-training corpus. Biases that, as I understand it, can never be fully overcome or patched with later training or guardrails, no matter how painstaking.

@danmcquillan @anildash
I suppose the "heat" parameter that determines "creativity, " --the randomness of which word/token is chosen from the ranked list of the next most likely-- could be set such that it's always the one at the top selected? That would still be Bullshit as a Service, I think we all agree? But is it anymore reasonable, just because it's the same old bullshit?