We need a new programming language suited for the post-agentic world of AI and the advent of the Machine Gods.That's Cosmos.>The entire base compiler of Cosmos is 100% written by a LLM (Astra).https://cosmos-language-lab-0819.asd-user56.chatgpt.site/Coupled with the qualities of readability plus maintainability (type and mode-checking). As a high-level relational GL5.5 (Generation Language 5.5th) programming language. To give you an idea, a language like Python is only a 4 in the high-level scale. A full relational language is minimum 5. Coupled with multi-paradigmatic capability, this makes Cosmos possibly the most high-level language ever created. All languages ultimately branch to Cosmos.Just compile and build to become a beta-tester. The language is already released for all coding hackers. (Some Gui-building utilities in progress).
Yeah buddy now go generate me some ai hentai so for me to goon and leave me alone
>>109984100Based. The future is unreal.
Hey JewGPT make me GL6 and make it better than this cosmos shilll, make no mistakes
>>109984100It should let app logic be modifiable in real time without having to restart the app. Live coding never really made sense at human coding speeds, but with AI it's a game changer.Clojure is apparently too difficult to work with, frontier models really struggle with it.
>>109984100"make the future, make no mistakes"
>>109984530there are no mistakes in computing. your program is just a reflection of you (retard)
>>109984530Pretty cool tagline tbqh
>>109984100Gay, I like my LLM-first language better.
>>109984100dogshit. rust is the vibecoding language, and you tell your house clanker the architecture.telling a clanker the actual code you want is just a massive skill issue and a red flag.
>>109986819>rust is the vibecoding languageTrue. if you make sure you have proper testing pyramid, including proptests and mutation testing so clanker stops writing vacuous tests it's pretty great. Haskell is also quite good, if you don't need the systems programming capabilities. And you can then have proofs in Agda and extract to Haskell. This stack is how I'm doing my LLM-first language and it works out great.
