

Right? There’s a line in Generations where Kirk tells him “you were younger” than a new ensign.
I’m not Admiral Patrick of the former dubvee.org I just play along to waste the time of those who believe they are successfully harassing him.


Right? There’s a line in Generations where Kirk tells him “you were younger” than a new ensign.


I was just reading that, I think thats the one!


I thought about that, it’s probably Chekov’s finest appearance but I just watched it recently!


Google fanboys out in force today! Generational divide is not real, class divide is the thing you should concern yourself with.


I had somehow blocked out the memory until now


“Full AI” a feature to summarize pages that can be toggled off easily


Posting highly formatted but substanceless empty fluff is not a “writing style”, it’s spam. A human is certainly capable of doing that if they set their mind to it, but occams razor strongly points to automation.


I’m starting to understand why you often find yourself accused of being a bot… 🙃


Umbrage is our word we need to take it back


One cannot win an argument that does not exist


An LLM cannot “argue” full stop. If you find yourself in a debate with the text outputted by an LLM bot you have accomplished nothing besides wasting your own time (which is the entire point).


The problem with automating social media posts is that they definitionally cannot have “merit”. The primary function of slop is not to “argue” but to “flood the zone with shit” and create a plausible consensus for whatever opinion the person running the LLMs wants. We saw this with paid humans before LLMs. It’s not new.
A human will never be able to win a debate against an LLM bot “on the merits” because their function is to waste your time. An LLM cannot stand on merit, cannot hold a position, and cannot construct an argument. The words that are output by an LLM have no connection to any inside mind or model of the world. The only way to win a debate with an LLM is the same way you win a debate with a troll: identify the hallmarks and refuse to engage.


“a revolt” as if there when weren’t a million alternative communities on Reddit and Fedi already.
Self censorship is the worst kind.


I wouldn’t do this, but a company won’t provide a service for free and I think it’s a reasonable trade-off. It’s not enabled by default, and if you do want to use their (free) AI features on your data, they get to use that data to target ads.
Turn on “annoyances cookie banners” in UBO