Siddharth Jaiswal

The year is 2025. Humans apply for jobs. AI decides who gets hired.

@nixxin Then you need opencode to work with free models 🙈

@sdrth I've hit limits in both. lol. Two apps over the weekend. A third in the works. I have about 8 app ideas so far.

Ok I've hit my Claude Code limit too, while building App #3. Very clear now that Codex not only has better limits (which I've exhausted), it also makes fewer errors. Anyone has tips regarding reducing errors and token usage while coding: please share. Also, tried antigravity again and it's still 🤮 Google really needs to up their AI coding game.

@nixxin Use direct models for interactive chats, make it write detailed task plan to markdown file (in-repo). Use Github Copilot to run a long session using this file, they bill per request. Loop this.

So @OpenAI Codex is getting a Kanban before Cursor and Claude? I just want @todoist board UI + quick entry UX to 100x Codex experience. CC: @thsottiaux

@_svs_ Wanted to use it with Hermes but both Swiggy and Zomato have whitelisted only claude,chatgpt, etc

Does the Swiggy MCP work? Anyone using it?

Agents need to evolve from being a Chat first UI. What is the next best UX for Agents?

We are back to the days of make money online with blogging using my affiliate links to blogging tools, hosting, wordpress themes, etc.

One of the hardest things I realise that every founder has to go through in their journey of building a venture is seeing - everyone else move on. Move on for opportunities, move on from situations, move on for their reasons, move on from conflicts, move on because they can move on. Everyone has that choice, not the founder. For founders to move on, they have to give-up. That's not how they are built, so they do all that is possible to win. Even if they win, or lose, either ways - they end up being lonely in their own journey.

Upward ladder in BLR HSR > Koramangala > Indiranagar > MG Road > JP Nagar > Nirvana Pls Note: Whitefield, Bellandur & Sarjapur aren't on the list. Intentionally.

When is @cursor_ai adding kanban view?

@dharmeshba x.com/sdrth/status/2…

Hostar has introduced vertical dramas on the mobile app.

Quibi was way ahead of its time. The market for vertical short web series is exploding 6 years later.

@sanketnadhani Try @p0, works well but needs cost optimisation Aside, @TheIshanGoswami retain this user

What's a good alternative to exa.ai? Another service that cannot process an Indian card.

Thousands of tools for PKM (personal knowledge management) but Obsidian + Git Repo >>> everything else.

@rohanvarma Automations, I had to turn off some of them because they used up a lot of my tokens.

If we made /slow mode in Codex, would you use it? What for? (Slower inference at a cheaper cost)

Confirmation from the last session. Codex actually sucks at UI tasks because it tries to demonstrate its thinking in the visual itself. It keep making those gradient cards with borders Example: an actual card with copy that says "this card remains sticky so users can have floating context available as they scroll" Claude has been doing great with UI and Gemini was surprisingly good at plain html css explorations. What is the training difference that's causing this? Note: Love Codex for everything else.

India still largely employs lift operators (for fully automatic lifts as well) in public places. The worst is having people stand next to a parking-ticket machine so they can just press the button and hand over the automatic ticket. Will AI need to solve for human-trust first in India? Probable pattern: human proxies. Humans delivering AI outputs on phone, in person, etc. Much like ticketing and self check-in kiosks where the product is designed for self-use but still employs a “human” to guide you through. Maybe self-driving cars with a person just sitting in the driving seat so riders can trust the autonomous system. Laggard’s take but still important for driving adoption.

Elevators used to be manually operated and were especially risky if the operator was tired or inebriated. Now, you just get in, press a button and a modern elevator, if inspected regularly, is extremely safe.

India still largely employs lift operators (for fully automatic lifts as well) in public places. The worst is having people stand next to a parking-ticket machine so they can just press the button and hand over the automatic ticket. Will AI need to solve for human-trust first in India? Probable pattern: human proxies. Humans delivering AI outputs on phone, in person, etc. Much like ticketing and self check-in kiosks where the product is designed for self-use but still employs a “human” to guide you through. Maybe self-driving cars with a person just sitting in the driving seat so riders can trust the autonomous system. Laggard’s take but still important for driving adoption.

@paraschopra Spot them in replies to this one

More than half of the replies on my tweets is from bots who simply rephrase what I just tweeted. What's the point of it? It's really sad.

too many agents running? you mean you're overwheLLMed