@byersblake@eddylong33 Yeah, because someone who’s been part of hundreds of deals would definitely only once casually suggest firing the founders and taking their equity. Total good guy behavior. Please keep defending him.
Not really laziness..spanish has a bunch of big latin american variants, and while the differences are usually not huge at a beginner level, they still exist. There are also national preferences and rivalries around vocabulary, accent, and “whose sspanish” gets treated as the default.
so Spain’s flag works as a more generic symbol for Spanish without picking a winner in Latin America.
people finding exceptions to prove a rule lol.... and that comparison doesnt really work, because Duolingo Spanish is not simply “Spain Spanish.” It mostly teaches a broad Latin American Spanish, while using the Spain flag as a generic symbol for the language. Portuguese is different because the course is specifically Brazilian Portuguese, not European Portuguese.
@robinebers cursor has the best harness by far for all frontier models, idk how ppl can make these comments still, it makes me think they are not really using all the available products or just farming/stupid
doubts? i’m starting to doubt anyone still defending anthropic’s products
the loudest claude code defenders seem to code less than 1hr/day, never hit the $200 limits, and somehow ignored for months that claude code had one of the worst harnesses while third party harnesses were getting better results from the same models
the models are good. the product around them is the issue
acting like claude code is ahead on usability, limits, and reliability feels insane if you use this stuff heavily....its just cringe
@dayonefoundry what you cannot vibecode is the harness and their data, claude code and codex still have not caught up, and cursor is still empirically best across frontier models