90% of coding is going to be fully solved in 2026

unfortunately being able to prompt well is still a skill issue other than vibe coding, the models still need decent instructions to get a good output luckily opus is a crazy good prompter, so 99% of my workflows is just asking opus to create the instructions for other agents

1/ Citadel got DeFi wrong. Today, DEF, @a16z, @DigitalChamber, @orca_so, @theblockprof, & @UniswapFND wrote to @SECGov in response to @citsecurities' letter misrepresenting how DeFi technology works. Why this group? Citadel blatantly miscited us, and we feel obligated to

I tried swapping from chatgpt to the Gemini app for a few weeks and wow it’s really awful I don’t know what’s going on but it’s way worse than just using the ai studio Barely listens to instructions, constantly have to repeat myself etc Not that chatgpt is any better with 5.2

most vibe coding platforms are still undervalued (even with all the crazy revenue growth) genuinely don't think anyone is ready for the influx of "software" that will be created in 2026 because of them. the masses haven't even heard of it yet ex: my wife wanted to build a mini

this is really neat. big fan of @dhruvamin and the team

crazy to say this but.. most of my day to day "general" ai usage (basic questions) has sorta moved to gemini i still use chatgpt app here and there (best app so far) but realizing i dont even use my 20/mo subscription that much

cursor would make so much money if they built a vibe coding platform

cope is unreal yeah models aren’t perfect but software engineering ain’t gonna be the same in 1-2 years its already different then 2 yrs ago swe role going to be less about writing code and more about reviewing /conducting and architecting

yeah anthropic was right, coding is probably done for pretty soon opus 4.5 + composer 1 + gemini 3 is basically an engineer at this point (backend/design/ iteration)

Man finally got around to using nano canna pro And it’s actually over I really wouldn’t believe any photo you see on online anymore

noticed a pretty worrying trend the more i use llms my day to day skills are slowly atrophying and im relying more and more on models for even simple tasks happening for coding, writing etc sometimes i dont even want to do the task if theres no ai to help

i know everyone says we're in an "ai bubble" but i feel like we're not even close? most people you talk to "tried chatgpt" (free btw) to write an email or summarize text/ content average business owner doesn't even know what an agent is or does

my cursor usage has basically 20'x with the new model i gotta say its going back to slow models suuuckks now

alright composer-1 inside cursor is reaalllly good like its definitely not the smartest model but its so fast that im using it for 70% of my requests now i think this shows just how important having speed is. all my requests are done instantly especially multi file edits

honestly notion is so terrible at ai i don't even bother touching it anymore just moved all my thoughts/notes/ to a single github repo with markdown and i'm getting way better results i can plug in cursor/cc/etc or paste into any chat app

just got my new mx master4 > click side toggle > interesting ai button > preloaded with chatgpt/perplexity/gemini/copliot wtf

probably the most important cursor launch ever they really needed their own model and now they delivered should be interesting see how good it actually is though

if you set chatgpt atlas as your default browser you get higher limits

chatgpt browser.. looks kinda good? rip to perplexity seems like direct competitor to their browser

the pricing for this is confusing, i don't really see anyone using this model over existing options