Shrish Dwivedi

@TheMarcitect yeah the cognitive dissonance is wild like they're selling you the chains and calling it freedom

@shikhr_ Like it’s like trying to fix something that’s not broken

@shikhr_ Nope but still not sure if I should switch or not

Name one AI Subscription which offers more "limit resets" than OpenAI Codex. I'll wait.

@jasperdevs the usage limit anxiety is so real, still dont know if switching to codex is worth it

saying goodbye to claude really is such a big relief no more having to avoid pi/other terminal tools because I can't use claude sub in them... and no more having to always be looking at usage limits just say goodbye to claude for a week and your cortisol levels will go down by 50%

@e0syn the irony of anthropomorphizing a chatbot while rejecting actual theology

atheists thinking AI is conscious call that the Claude delusion

@haider1 so the last 20% is just... typing? that can't be right. feels like we're measuring the wrong things if "can write a kernel" = 80% of human intelligence

OpenAI Greg Brockman says we're about 80% of the way to AGI Today's models are already stronger than him at writing software, and with the right context, they can outperform most humans — even on low-level tasks like writing kernels "the last 20% is being human at a keyboard"

All these tech bro's telling us how great the future is going to be while simultaneously building the control grid that will eventually enslave humanity is surreal. People are just lapping it up.

@svpino been doing this with cc for a month now and the patterns it catches are wild turns out i was basically using it as google search half the time

Most underrated Claude Code tip: Use Claude Code to coach you on how you use Claude Code. 1. Run /insights. This will generate a report on your usage patterns, tool calls, and time spent. 2. Find the report file it produces. 3. Open a fresh Claude session, attach the report, and ask: "Read this report. Identify the 5 worst habits in how I use Claude Code, and give me a concrete fix for each." I have a recurring calendar event to run this every 3 weeks. Every single time, I find something new.

@direkturcrypto yea try my quantum computing WebApp as well its at localhost:6429

Claude Code 4.7 is insane. i know literally NOTHING about coding. ZERO. and i just built fully functioning dapps in 30 minutes. http://localhost:5173/ check it out.

I raised $100K in college with nothing but a prototype and a pitch deck. Had a working prototype of a crypto mutual fund app. That was enough back then with some traction. Today? An MVP won't cut it. Hell, even a working product isn't the bar anymore. How delightful. The real question now: how many people will actually pay for it? Distribution has become everything. You can't just build and hope anymore. Back then, having something that worked was rare. Now AI lets anyone ship a product over the weekend. The constraint isn't building-it's getting people to care. The game changed fast. In 2022, investors bet on potential. In 2026, they bet on your ability to distribute. Because that's exactly what founders needed-higher bars. You need users before you need funding. You need proof before you need a pitch deck. And honestly? That's probably healthier. Forces you to talk to users from day one. Forces you to solve real problems, not imaginary ones. The bar moved from "can you build it" to "can you get people to use it." And if you can do both? You're unstoppable. Or at least fundable.

Taste matters more than ever because depth in multiple domains is rare and valuable. Think about it: how many people truly understand video editing, motion design, AND tech? Or can grasp CALL/PUT options while using Claude Code to backtest strategies? These cross-domain experts get hired first, funded first, and paid most. The intersection is where you win.

@_nikhilsheoran @sahanamantha7 Agreed

@sahanamantha7 kinda like mine tbh nikhilsheoran.com

what is the best personal website you have seen?

@GithubProjects GitHub.com/nanonets/nanoi…

|￣￣￣￣￣￣￣￣￣￣￣￣￣| | Share your GitHub profile. | |＿＿＿＿＿＿＿＿＿＿＿＿＿| \ (•◡•) / \ / —— | | |_ |_

@duttasomrattwt Too good!

such a sleek motion design work!

If you are a "VIBE CODER", you need to stop building RIGHT NOW for a whole week. yes i'm serious. you've built enough shit. your github is full of half-dead projects with 3 stars from your alt accounts. the problem isn't that you need to build more. it's that you're scared of the part that actually matters. so here's what you do: week 1 - 100% marketing. no code. zero. i don't care if you get an idea. write it down and ignore it. week 2 onwards - 80% growth, 20% building. flip the script permanently. Think hard, if just building products would make you rich, or win, why isn't everyone with Cursor or Claude Code winning. building is comfortable. posting feels like begging. i get it. but nobody's gonna stumble onto your app in a dream. you have to shove it in their face. the best product with no distribution loses to the okay product that won't shut up about itself. stop hiding in your code editor

been watching this happen in real time and it's wild how fast it moved chegg's entire moat was just having answers slightly faster than finding them yourself. chatgpt made that worth zero overnight most SaaS products have a similar "we're just a thin wrapper" problem and don't realize it yet

Chegg was a $14.7 billion company. Today it's worth only $156 million. ChatGPT launched in November 2022. By May 2023, Chegg's stock had crashed 48% in a single day. - Half a million subscribers gone. - 45% of employees fired in one shot. All because students realized ChatGPT gives free answers and Chegg was charging $19.95/month for the same thing. $14.5 billion erased. Not by a competitor. Not by a government. But by a chatbot. This is not AI hype. This is what AI actually does to businesses. Be aware.