infosex.exchange <3
You are probably looking for the infosec.exchange Mastodon instance
This host is mostly for my random stuff, and in little part acts like a well-intentioned placeholder for the typosquatted domain.
Discoverability and Archiving
Currently I'm using this host for saving the items from my own feeds to the Wayback Machine and provide in-links for search engines. I hate that I have to do this, but the non-sense ideology of Mastodon pretty much ruined the search feature for Fediverse as a whole, and this wasn't changed by the fact that they owned their mistake and implemented search eventually.
Yes, I (or anyone else) could do similar things with other peoples published feeds, regardless of the tantrum. No, you can't defederate this, because the process doesn't rely on an instance.
Gluttony Section for Search Engines
@Viss Wasn't that X-Files and a large underground mushroom colony?
@freddy But seriously, I just added a comment to my query and I swear it got slower...
@freddy Is that even a requirement these days?
on the back of the envelope, counting with an avg. yearly salary of $75k for a teacher in the US, the projected $4.8 trillion AI market by 2033 would equal ~7M years of teacher salary every year.
#weirdunits
@d_olex Yeah I get that. My point is (but I'm unsure about history here) that when Java or first browser JS engines were shipped inefficient solutions were probably necessary, and now we try to reduce that debt, while in case of your modern examples we probably have cheaper solutions that work better, but burning GPUs is sexier.
@d_olex Good question, but I'd argue that bytecode solves existing problems, while in case of LLM/blockchain I mostly don't see that. Also, isn't JIT specifically a thing to improve performance, meaning less resource consumption? A related observation is that many use-cases for LLMs can probably be solved much cheaper, today. E.g.: better IDE features; more QA for web search results; better education so people can write and understand an email.
@kagihq "I don’t think I need to list the large number of tasks where LLMs can save humans time" - I naively thought this would be the whole point of the post? It'd be also important to back up that "large number of tasks" with data (e.g. time to result with/without LLM).
Next Page