nclin_

retention for 'safety' -> AI race as national security -> training on your data for 'national security' aka safety

It's simple mental calisthenics. If you are handing an organization whose entire business model is built on stealing data with spurious reasoning, what do you actually expect they will do? Don't be a fool.

Aahahahaha yes the solidarity of the common memecoiner must not be broken.

Is claude code the best coding harness? Anyone running evals on that?

What are the VSCode features you find difficult to replace? Zed is quite good as a similar workspace, but I can't speak to extensions.

It's your own fault for using anything made by microsoft at this point.

Ah, to have "kill a million people and don't go to jail" money.

The last few days I've seen more degradations and canceled my Max subscription.

Presumptuous and wrong "memories" from a one-off command which affect all future commands, repeated/nonsensical phrases in messages, novel display bugs which make going back in the conversation impossible (I can't tell where I am), lack of basic forking features (resume a current convo in a second CC instance -> fork = no history for that convo?), poor/unclear reasoning, a new set of unclear folksy phrases (it really wants to "cut code" all of a sudden).

Qwen + Opencode has been a game changer: which runs very well on a 4090 for basic/exploratory/private tasks, and being able to switch to and between frontier models (using openrouter in my case) to avoid vendor lock in feels like basic hygiene.

There's also the homo economicus psychological difference between having a token budget to use up, and a cost per token. I'm more thoughtful about my usage now.

He was directly addressing apps that claim to do so, and proving that they can't, with laypeople who might have diabetes as the target audience.

It's more a data-driven pub test, I think it explains itself well.

The question is - are those apps actually so simplistic? Or is this a strawman.

Construction regulation is the preventative measure.

Pillage libraries for LLM training data, then sue them and shut them down for archiving a different format. Cool.

Proposing a definition of slop: content optimized for profitability, regardless of quality.

If AI slop is replacing the content you were consuming, it was already slop.

Meanwhile, Sora and other video models have become available on Openrouter. Are we sure this is the end of Sora? Or just the interface?

weird, still dead here (entire rest of the internet is fine) :shrug:

Dead link