@trq212@keyurrao there are two types of problems: incentives and culture. incentives are labs shipping faster because profit demands it. culture is how people choose to live with what ships. what you're describing is a culture problem
@jp54362 Astra is the best model Iโve ever used. I was a hard-core Claude code user for the past year, but I canceled my Anthropic subs. Sol was a decent holdover, but finally, OpenAI released something significantly better than Fable (itโs cheaper and faster), plus better computer use
@Greg_GL_87@iuditg fable isn't even that good with the safeguards - we can't use it for debugging because "cybersecurity". Their output has degraded significantly lately
@ali_aka_ahmed@UseCorgi Cyber insurance for enterprise contracts. Go with https://t.co/E8wTiV3tx7 though, corgi is taking risk of AI startups and basking it togetherโฆ who is underwriting that???
@seek2become@Sunny_and_72 Perhaps, but this creates massive downward pressure on token prices, especially once memory chip supply catches up. Iโd love to run more locally for my business