MrCheeze
banner
mrcheeze.github.io
MrCheeze
@mrcheeze.github.io
I reverse engineer games, etc. Also found elsewhere on the internet: https://mrcheeze.github.io/
January 11, 2026 at 1:54 PM
First? Those kinds of subreddits are widely understood by their users to have plenty of fake highly upvotrd stories, including AI written ones
January 6, 2026 at 8:23 PM
For You is better but there's still far, far too much uspol just for the resson that it is unfortunately most of the posting on this site
January 5, 2026 at 10:05 PM
Unironically might have killed the site
January 5, 2026 at 9:40 PM
Stark, but much less stark than the decline for programming questions - I assume because MathOverflow is actually used by mathematicians for original research whereas StackOverflow is strictly a repository of what's already known?
January 5, 2026 at 9:05 PM
It does happen with CharacterAI, which has low market share compared to the big 3 LLMs but targets engagement above all else
January 5, 2026 at 8:52 PM
Threads is to this day the only twitterlike with a population of ~Normal People! And yet for that reason itself it is largely ignored by us overly online
January 5, 2026 at 8:40 PM
draws the word "poli" on the chalkboard. audience frowns. connects the "i" into "l" and the audience bursts into applause
January 2, 2026 at 6:46 PM
bsky.app/profile/did:... is what most power users use nowadays instead of Discover, but there's still a ton of polislop in there, in stark contrast to Threads and old-Twitter which were actually good at serving posts relevant to my interests (even when I'm trying to avoid them, it seems)
January 2, 2026 at 5:19 PM
is india a continent Y/N
December 31, 2025 at 2:12 PM
That's why it's not called friends and acquaintances things
December 30, 2025 at 2:54 PM
Dune 2 chicks at the same time
December 29, 2025 at 4:53 PM
I've seen people describe using them as "gacha", so the idea is definitely out there at least
December 29, 2025 at 4:53 AM
I was going to say that it's probably just inattentive to the fact that such a framework even exists, but if even the first reminder didn't work, then I'm probably wrong about that explanation
December 28, 2025 at 6:04 PM
To be exact, the human instructed the agents to do random acts of kindness, and Claude's idea based on that was to email people

(Others did other things, GPT made a rather useless Google form)
December 27, 2025 at 1:12 AM
They could suggest it to new users but the raw numbers aren't high enough to justify doing so, unless you just consider it to be manual curation, at which point the decentralization is lost
December 26, 2025 at 4:09 PM
For sure but "liked by 22k" is still a very small people and means most users (especially new users) are NOT getting the good algorithm and therefore are likely to just bounce off the site
December 26, 2025 at 4:09 PM