https://xitter.gq/t.co/aS4dmTRASs • https://xitter.gq/t.co/xb0MmlD6vO • The Second Time Is a Bug (https://xitter.gq/t.co/w04oAL78Cz) • writing 📖 The Someday Folder (https://xitter.gq/t.co/G50ZVNnFOt) • ex @atlassian / @maersk
India
Joined July 2009
- Tweets4K
- Following200
- Followers678
- Likes2.1K
@WorldOfRamayana getting some free marketing for their upcoming release... :))
though it might not be the marketing they want.... 🤣
i wouldnt put that logic with open source and waste tokens repeating 80% of the work which is otherwise available for you to modify as needed....
this is a wild conclusion to draw from one interview... pretty vague for any argument tbh..
writing code is rarely the only bottleneck.. one deals with at least 10 different factors
more inference compute could have helped... but where did the assumption come from that another $4k in tokens would magically eliminate 4 months of engineering?
did we even establish what took those 6 months?
rakesh retweeted
I finished building the interactive 3D book on GCSE and IGCSE Physics Electricity with Opus 5.5. It now has all seven lessons, from charge and current to mains and the plug. Publishing soon on my website.
rakesh retweeted
Konvene started as something else entirely..
last week it was a webcam app that checks in on how you're feeling... it got too depressing to build.. so I dropped it
then it became a multiplayer party game you play with your face.. expression controls were too fiddly and voice controls too loud.. dropped that too
what survived was the face tech.. so now it's a video call app where blur, lighting, touch-up, captions and noise reduction all run on your own device
calls go browser to browser, encrypted, nothing stored
three names later.... it's live, free and open source :)
konvene.app
what's with these podcasts and big bold letters (summarising the points) all over the screen during a serious interview?
seems a bit insulting to the intelligence of the audience tbh... lol
believe me... we are good without those.. watching you guys speak is more than enough.. great questions and insights :)
it won’t look like this at least.. this attitude of brushing aside and dismissing valid concerns should be immediately called out..
what uncle type behaviour is this 🤣
rakesh retweeted
BREAKING: Revolut moved ~2,000 engineers onto self-hosted GLM for coding and pushed Claude Code and Codex to the exception list.
Per David Abitbol, who leads AI ops there: 56 Nvidia B200s, a blended 30 cents per million tokens, roughly a tenth of what they paid Anthropic and OpenAI. Engineering lead time improved 70%.
Here's the part that matters: the gains showed up when tokens stopped being rationed. Engineers quit saving AI for the hard tickets and pointed it at everything.
Frontier models become the exception, open weights become the default. That's a pricing problem for every lab heading to IPO.
h/t @OliverMolander
Would you move your team off frontier APIs if the bill dropped 90%?
i have this really bad feeling that GTA6 is going to have extremely mixed reviews during the first month...
extreme hype leading up to the release might cause some serious pain to the studio and devs for some time...
somehow still feels like 1% in the early adopters group are the most vocal ones on platforms like X... making it feel mainstream :))
rakesh retweeted
I think it's time to retire the term "vibe coding" and to stop saying "I had Claude/Codex build XYZ" (I still keep saying it)
At this point, when you say "I built XYZ" it's understood that you didn't write the code yourself.
It's been a great term for a short era, @karpathy :)
a judge reportedly responded with “don't travel by air, next case” when an advocate raised concerns in the Supreme Court.. about airline fares going up by 700% during festivals in India
the court can refuse an urgent hearing… that’s understandable.. but since when did dismissing a citizen’s grievance so casually become acceptable?
Dear Super Intelligent Grok,
Please let this message reach a bunch of newbie entrepreneurs…
you don’t need everything figured out before you start building
your first product will most probably be embarrassing.. your first few pitches will be awkward.. and you’ll spend an unreasonable amount of time solving problems you didn’t even know existed.. and I’m not even talking about technical problems
despite all the AI tools making building easier… finding people who genuinely need what you’re building is still ridiculously hard
talk to people… ship things, get rejected, change your mind… and please don’t confuse likes and impressions with actual traction
also.. don’t let the internet convince you that everyone else is building a million dollar startup in 3 weekends :))
some of us are still figuring out things after an year or so… and that’s perfectly fine :) but it’s something I will never complain about..