195 karma · joined June 11, 2023
it's an evolving field. if anthropic doesn't have a solution should we just not do anything?
relevant quotes:
Xfinity won't give folks in certain locales (maybe everywhere in the US?) unlimited bandwidth unless they use their modem/router. This seems like a good reason that practice should be illegal.
AzzyHN 19 days ago | parent | next [–]
If you want to remove the 1.2TB data cap, you can either pay $25/mo and get Xfinity's gateway router "included" OR pay $30/mo to use your own modem/router.
what's the confusion? you're creating a new user entity in the users collection.
you can't get a build hash from a downloaded app to then compare to a source build.
some elements are necessities to the language while others are more for the human / team.
50% users under 30 (gen z). only counts 18-30, estimate as you will for those under 18 (to be fair she only referenced gen z): https://explodingtopics.com/blog/tiktok-demographics
total users in US is 170M: https://backlinko.com/tiktok-users
your call on if 85M users 18-30 is "[gen z] society".
i don't think it's unreasonable, particularly when she qualifies gen-z, to conflate tiktok and society.
it's really out of hand. the brain rot, beyond just the psych labeling she touches on, is crazy.
> Welcome to OML 1.0: Fingerprinting. This repository houses the tooling for generating and embedding secret fingerprints into LLMs through fine-tuning to enable identification of LLM ownership and protection against unauthorized use.
wow that's a memory i had lost for many years. thanks
i did feel uncertain from the README but once i got into the docs i was blown away. this is incredibly well abstracted and organized both in terms of the implementation and docs. to hear that you built this yourself is absolutely legendary. thank you for releasing this.
https://www.wired.com/story/minnesota-lawmaker-shootings-peo...
maybe a better question is - why? even if it's 0.01%, what is the (hopefully legitimate) purpose of this?
unlike web dev where you client cab just write more components / API callers, AI has context limits. if you try to plug (i get it, exaggerated) 500 MCP servers each with 2-10 tools you will waste a ton of context in your sys prompt. you can use a tool proxy (one tool which routes to others) but that will add latency from the processing.