Nitesh

@nileshtrivedi @NandanNilekani I don't know why anyone would even think that small. Everyday I get at least 5 thoughts on how I am gonna figure out AGI before US labs😂... I mean at least aim high bruh.

I strongly disagree with this. And I cannot believe how @NandanNilekani can be so short-sighted. Data and intelligence sovereignty should be a national strategic priority. Using AI models from foreign players will involve perpetually sending them all of our data, including use-case insights and user preferences.

@NandanNilekani @rvenk @EconomicTimes O bhai rehne de. Retirement ka time aa gya h. Tu FD kara and mauj kar. Leading the world baaki log kar lenge.

India doesn't need to lead the world in building the most advanced AI models. But it must lead in ensuring benefits of AI are widely shared. @rvenk and I have an op-ed in The @EconomicTimes economictimes.indiatimes.com/opinion/et-com…

@abhishekcode42 bhai 11th class mein pahuncha dia lol

Chat, let's solve this in 5 seconds

@sporadica Yeah...its a bit too ambitious. There needs be around 10million people on moon for 1m to reach mars. In his lifetime I think a lunar base looks more probable.

trying to genuinely understand this bc this feels...impossible in his lifetime? $7.5T is crazy, but okay, doable. but 1 MILLION people on Mars?? Yea I believe we will get there eventually, but Elon is 54yrs old Either he thinks he's gonna live to 200 or we'll do this in ~30yrs?

@archiexzzz Someone please tell the infosys guy about power law. 3B$ leke kaha jaenge ye log. This is a fuck ton of money by indian puchase parity. Should just invest every last dime.

Huge respect to Nilekani Ji, but I disagree with him here. Don't take this advice. It's outdated and has bad taste. "India does not need to" WHAT? Imagine if Homi Bhabha had said this about India's Nuclear Program. What BS is this? India built a full nuclear program from scratch, under sanctions, with zero help from anyone. Nobody gifted it to us. If we can build nuclear weapons, we sure can build SOTA models to compete with China and the USA. Why have this "shoot for the 4th floor" mindset while others are shooting for the moon? India has everything: the talent and the brain power to compete with global players. So what is the constraint? Capital? Why? Because Infosys, with $3 B in net profit, didn't invest a dime in R&D - because they are "too afraid" and want to play it safe?

A very limiting thing for the current AI is that its very good at what people were doing a lot of...markdown, typescript etc. But its very hard to make it an expert at something new without a humongous training data. This tells me there isn't an element of "intelligence" here.

@rezoundous If anthropic fixes its usage...then no. Else definitely they're gonna start the crunch too

What are the chances OpenAI reduces Codex usage limits in May?

@Hiteshdotcom @github Apna college is anough to take down github😂

It’s exceptional pressure on @github Every vibe code app, vibe coding platform is putting up an insane pressure on Github infrastructure. I am just surprised that they are not charging these companies. They happily pay to Anthropic and OpenAI, they can pay for GitHub infrastructure too. Great company, great product but pressure is too much at this point. All AI flood is at one direction, GitHub. I guess, this is a big decision time for them. And decision is such a critical thing that no matter what you do, there will be enough people to dislike it or rant about it.

@RhysSullivan We was hoping AI would have some idea

we literally have no idea how to make software anymore

@peer_rich @pumfleet

the techbro urge to build a github competitor that is able to promote open source without its downsides would be an absolute chad character arc for @pumfleet and me hahahah

@_CallMeMacy The only issue is that you must resist the urge to do too much with it. It will make it complex and then maintenance will become a headache and even claude wont be able to help you.

Lately, I've been advising founders to vibecode their own CRM that reads their emails, calendars, and keeps things up to date in real time. It took me 2 hours to create my own using Claude Code (and I've never coded in my life). Am I crazy for suggesting this or do others agree?

@shantanugoel Dario ko shovel bechne h aur nikhil somehow mast vella h. Dono ko baithke badhiya bakwas karne ko mil gayi or kya chshiye inhe. And as it turns out public bhi mast velli h inki baate sun rhe h sab ke sab🤣

Dario/Kamath podcast is going viral again. I don't know why people listen to either of these people. One has a single rhetoric since last few years: SW Engineering will be dead in 6 months The other is a glorified pre-prepared question-asker with nary a dialogue to eek out anything worthwhile. He should take classes from Dwarkesh/Jensen.

@athyuttamre Now we have 100x faster bug. Its time to get to work.

software engineering got 100x faster and i don't know what to make of it

@madewithmiso link? Also why have a name and a handle? Just have the handle and remove the @.

twitbruv clears X and bluesky. we can even play chess on there.

@danizeres Which is this tweet. Bravo.

content about codex

@EigenGender He realized no more mega funding rounds coming and he'll need the money from humans.

kinda weird how Sam Altman no longer believes in agi

@kunchenguid toolception is real

i ask agent to build a thing i realize a tool can make that easier i ask agent to build the tool i realize another tool can make building tools easier i ask agent to build the other tool i now have 10 tools. still haven’t built the thing help

gpt or claude? Guys we need both of them. If either one is gone then the one that's left will eat us.

@leerob The most impact is caused by making people think less. Which is counter intuitive cause the whole points of agents writing code is so you can do the actual thinking. But the thing that makes humans think less and work lazily will take off. (It'll be important to not fall for it)

It wasn’t obvious to me one year ago that an excellent coding agent would also be the path to a general agent for all knowledge work. But now it makes a lot of sense. I’m interested to see where AI is at next year and what seems obvious then in retrospect.

This is literally it.