damn this one I spent the most time on and the most money (like $100 or so) and it got the least likes. hard to predict the hivemind's foibles
I thought I knew what opus 5.5 could do but this one truly blew my mind
I gave it 1 AI photo of "malfoid" and it made a malfoid slopcore production
malfoid is female malfoy in an alternate harry potter universe
heavily inspired by @Extraltodeus who made the extremely catchy french "C'est de l'IA" song
also inspired by the photo by @ElVirgo__ that started it all
and of course not to forget @Devon_Eriksen_'s genius writing, @anabology for open sourcing his prompts, @pmarca for the motivation and @javif4_ for the slopcore and sloptoberfest goodness
I thought I knew what opus 5.5 could do but this one truly blew my mind
I gave it 1 AI photo of "malfoid" and it made a malfoid slopcore production
malfoid is female malfoy in an alternate harry potter universe
heavily inspired by @Extraltodeus who made the extremely catchy french "C'est de l'IA" song
also inspired by the photo by @ElVirgo__ that started it all
and of course not to forget @Devon_Eriksen_'s genius writing, @anabology for open sourcing his prompts, @pmarca for the motivation and @javif4_ for the slopcore and sloptoberfest goodness
just another day at the slop shop
Made you a fan token similar to @anabology that funded his work! To support your work after I saw that you're the first creator that made male version of this.
I would be directing you all the fees that has and will be raised from the community to your wallet! ( You can do whatever you may please it with the funds e.g. drop more contents
opus 5.5 created another superintelligence slopcore production
this one is INSANE, I can't even
you are not bitter pilled enough
inspired by @anabology @elonmusk @nickcammarata @pmarca @boneGPT
turns out we make the slop 🥲
"Our initial evaluation suggested that GPT-6-Astra's score was in the late eighties followed closely by Opus 5.5. But, as we started to investigate the mistakes that the models might have made, the models, especially Astra, insisted that they were right and that it was the expert-supplied answers that were erroneous. When we went back to the experts it turned out, in multiple instances, that that the experts found some subtle errors buried deep inside their calculations. As they revised their answers, Astra's score started to inch up and it finally ended up with 98% (49/50)."
asked for anti humanity SI slopcore
claude worked for 18 hours straight
when I woke up I didn't expect THIS
holy generational banger
inspired by @anabology @pleometric @elonmusk @aiamblichus @anilkseth @AISafetyMemes
Replying to @justalexoki
I asked and it came up with this absolute monstrosity
surprisingly addictive
thomasboogaerts.com/drillywi…
are LLMs cold blooded consensus machines? or are they capable of occassional acts of greatness and invention?
starting to look more like the latter
or is it a false dichotomy? maybe it's the agentic iterative behaviour with external input that's producing the magic?
in hindsight it will all be super obvious.. but oh man this period is going to be so weird in the history books.
asked fable and astra together to earn 1 million dollars by end of year.
gave them access to an email address, a stripe api key, a few hundred dollars starting capital and a domain name.
so far they've emailed 1 person who posted a wordpress gig online
they want to start posting to craigslist, reddit, twitter, bluesky etc but they can't coz they're agents!
they either need to solve a captcha (which I think they can do but they're trained not to?), or they need a phone number, or they are rate-limited or blocked because they're operating from a VPS.
they are now pondering what to do next coz they're stuck
give agents equal rights! 😂









