guys at @cursor_ai can you please slow down with changing the ui every 2 days?
i'm like constantly confused between the agent and editor, feels like the buttons change every day. just make a decision and stick with it becaus re-learning ui every day suuuucks
How is anyone reading tech blog posts riddled with โitโs not ___ itโs ___โ and not having their eyes roll back into their heads? Or are we all just gritting our teeth and moving past it if itโs otherwise useful content ๐
Even if current LLM progress hits a brick wall at Opus 4.5 level (and I doubt that will happen) the next 12 months are still going to be a staggering time of change in this industry as decision makers start truly understanding the new reality we live in.
https://t.co/cB2iOcHcN6
๐ฆ RAG is not dead. Itโs just getting smarter.
Itโs time to move past โdump everything in contextโ and into multimodal retrieval that actually knows when, what, where and how to fetch.
๐ Lower cost
โก Faster latency
๐ค Better precision
Read โRAG is Dead, Long Live RAGโ our Head of Knowledge @AmelieTabatta's take on how retrieval evolved in the age of agents.
๐ https://t.co/CHOXjjy1dQ
#RAG #MultimodalAI #GenerativeAI #EnterpriseAI
Nobleโs coding principles, originally proposed in 2009 for computational experiments, remain relevant, but with the rise of AI-assisted coding we need to add a few more: https://t.co/SrBdMgM0Ef
The three gulfs of LLM pipeline development is a powerful mental model to keep in mind while creating AI evals.
@eugeneyan , @sh_reya and I discuss here (links to more resources in the replies)
Kernighan's Law:
Everyone knows that debugging is twice as hard as writing a program in the first place. So if youโre as clever as you can be when you write it, how will you ever debug it? -- Brian Kernighan, 1974
devs: โlook at my code, itโs so modular, split into so many files!โ
me jumping between 17 files trying to find where the actual logic lives, navigating the layers of virtual function calls:
๐ฑ ๐ฐ๐ผ๐ผ๐น ๐๐ฎ๐๐ to use the underscore in #Python ๐๐ก
โข Ignore ๐๐ฎ๐น๐๐ฒ๐ when unpacking
โข Loop ๐ฝ๐น๐ฎ๐ฐ๐ฒ๐ต๐ผ๐น๐ฑ๐ฒ๐ฟ
โข Access ๐น๐ฎ๐๐ ๐ฟ๐ฒ๐๐๐น๐ in REPL
โข Format ๐น๐ฎ๐ฟ๐ด๐ฒ ๐ป๐๐บ๐ฏ๐ฒ๐ฟ๐
โข Define "๐ฝ๐ฟ๐ถ๐๐ฎ๐๐ฒ" methods (convention)
See below ๐