Posts by kopper@not-brain.d.on-t.work
 (DIR) Post #B6lEqVkJH9hnsBPjxQ by kopper@not-brain.d.on-t.work
       0 likes, 0 repeats
       
       every time i come across yet another another ai generated image i keep thinking to myself some perv on civitai couldve made a model that generated a better looking job and gave whatever's on screen huge bazonkers
       
 (DIR) Post #B6lEqVs6o9waGN3y76 by kopper@not-brain.d.on-t.work
       1 likes, 0 repeats
       
       is this enough? did this lower the rent on this account yet
       
 (DIR) Post #B6lEqW2kAcS0nM2Sgq by kopper@not-brain.d.on-t.work
       2 likes, 1 repeats
       
       if you're an "ai artist" and your tool of choice is like sora then sorry you're just a sucker
       
 (DIR) Post #B6lSJpFr7W3aBFThk8 by kopper@not-brain.d.on-t.work
       1 likes, 0 repeats
       
       @fiore that's a custom tool that re-prints the parsed AST
       
 (DIR) Post #B6ln4QAIDdx13lN7Hk by kopper@not-brain.d.on-t.work
       0 likes, 1 repeats
       
       ext4 lost+found folder but it actually p2p synchronizes random lost media for you to find
       
 (DIR) Post #B6lwJXDlAOoKsHH7PE by kopper@not-brain.d.on-t.work
       2 likes, 0 repeats
       
       look at him pretending like he has any idea what that ball does. you've never balled in your life. fake ass fan with your fake ass countdown fuck you
       
 (DIR) Post #B6nRaW9xenwAF82JVY by kopper@not-brain.d.on-t.work
       0 likes, 0 repeats
       
       @david gonna be honest this feels unnervingly LLM generated. i'd appreciate if someone verifies the sources but sentences like "The topic-matching confidence threshold is 0.8" (0.8 of WHAT, exactly?) does not really inspire confidence
       
 (DIR) Post #B6nSLXDHYddT7Z0ct6 by kopper@not-brain.d.on-t.work
       1 likes, 0 repeats
       
       @natty @david it's in one of the dropdowns at the end, ctrl+f would've worked if it was a details but that's what we get when LLMs get involved
       
 (DIR) Post #B6nloFMgtVWHICoYcK by kopper@not-brain.d.on-t.work
       0 likes, 0 repeats
       
       is there any existing knowledge around caching response "fragments" individually or Just Joining All The Tables? all the search engines suck nowlike, if i'm displaying a list of Posts in a web api, wouldnt it make more sense to render a Post once and cache the rendered JSON? the tradeoff here is that you can't Just Join the "whatever_list" table to the "posts" table (and to the "users" table, because your user data would also be cached in this way)of course you'd still batch the queries to avoid n+1
       
 (DIR) Post #B6pLJzVx3Lv2x9yiKe by kopper@not-brain.d.on-t.work
       0 likes, 0 repeats
       
       is there literally anything fun anymore or is that just me
       
 (DIR) Post #B6qtdLg8SiggnQ6UBE by kopper@not-brain.d.on-t.work
       1 likes, 0 repeats
       
       nocoffei.com/?p=451
       
 (DIR) Post #B6t6Hw587gkT31pn3g by kopper@not-brain.d.on-t.work
       2 likes, 1 repeats
       
       "small details in my mastodon client that i wanted more people to notice"w.on-t.work/outpost-frontend-details
       
 (DIR) Post #B6zW6mw7IGsTDjWX0C by kopper@not-brain.d.on-t.work
       0 likes, 0 repeats
       
       unpolished thought: a rough equivalent of the "followed by people you follow" signal for cases where you can't determine the exact set of someone's followers for privacy, where you analyze their replies to / from people and check if they have frequent back-and-forths (and maybe other signals such as likes and boosts and quotes) to approximate if they're aware of each other enough to be listed as such
       
 (DIR) Post #B6zfgbKAqLXnDdvaqG by kopper@not-brain.d.on-t.work
       1 likes, 0 repeats
       
       I can easily find the ytp uploads even without thumbnails by just searhcing from:kopper has:video and finding the ones that nobody interacted with
       
 (DIR) Post #B7O5Sh4ZDrShr3hTEm by kopper@not-brain.d.on-t.work
       0 likes, 0 repeats
       
       value removed tax
       
 (DIR) Post #B7O5Y58hAfDlwaCU2i by kopper@not-brain.d.on-t.work
       0 likes, 0 repeats
       
       update: i ruined it more
       
 (DIR) Post #B7O5Y5HucOasPAVqPQ by kopper@not-brain.d.on-t.work
       0 likes, 0 repeats
       
       i really need a larger cleaner corpus to test this in but idk where to find one lmao
       
 (DIR) Post #B7OiVLCCueU15PWjTM by kopper@not-brain.d.on-t.work
       1 likes, 0 repeats
       
       due to ai scrapers, we now require visitors to fold  one (1) protein before entering this website. thank you for your understanding (processor fans ramp up)
       
 (DIR) Post #B7Q4XlpZFSFQQCSdqi by kopper@not-brain.d.on-t.work
       1 likes, 0 repeats
       
       in french, they use the le prefix to explicitly indicate the thing they're talking about is in little endian.in greentexts, they use the be pr
       
 (DIR) Post #B7R5ILHrwrcOrnrkUy by kopper@not-brain.d.on-t.work
       1 likes, 0 repeats
       
       github.com/whatwg/html/issues/12584