Baldur BjarnasonMar 22

“Context Widows - by Kevin Baker - Artificial Bureaucracy”

https://artificialbureaucracy.substack.com/p/context-widows

> Goal displacement is a different diagnosis that is being made on a different patient. The problem is not in the metric but in the organizational form that needs metrics to function.

Context Widows

or, of GPUs, LPUs, and Goal Displacement

Artificial Bureaucracy
Show threadBaldur Bjarnason

https://artificialbureaucracy.substack.com/p/context-widows

> They did not arrive as disruptors. They arrived as intensifiers. LLMs function as an accelerant for the existing optimization machine, making the logic run faster rather than challenging its foundations.

This is effectively a continuation of Joseph Weizenbaum’s decades old argument that computing encodes and optimises existing structures. LLMs go further: encode, optimise, then intensify.

Context Widows

or, of GPUs, LPUs, and Goal Displacement

Artificial Bureaucracy
Mar 22 at 11:11amWeb
Show threadRuud SteltenpoolMar 22
@baldur Is https://communities.surf.nl/en/regie-op-data/article/fair-data-too-little-too-late-unless challenging some of its foundations?
FAIR data: 'too little, too late', unless …

When people say "make data FAIR" I wonder whether they mean doing extra paperwork after the fact, or changing how data are created so they’re FAIR from the start —

SURF Communities
Show threadBaldur BjarnasonMar 22
@steltenpower No. Seems to be the product of an entirely different conversation with no bearing that I can see on the essay I linked to.