Shrinath V

Hey @ICICIBank, you're going to charge 3.5% markup fee now ? For what joy ? Every claude subscription I pay for now gets costlier. I need to shift to another card if so

Claude code mods with some adjustments is turning out to be my favourite feature.

I can't believe @github still doesn't have pinch zoom for images on the mobile. Trips me up every time.

anyone have a spare invite for underdog? @underdogai

Why is no one penalized for the jams in Gunjur-Varthur? Every parent I speak to is mad - buses have been stuck for hours on the road. Absolute mismanagement - @krishnabgowda @GBA_office. The work was supposed to be completed by 4 Sep. No contractor or GBA official is penalized for delays. Everyone is helpless - and no authority is bothered.

I wish we'd used "fabrication" instead of "hallucination" when talking of why LLMs make up stuff they don't know. Hallucination implies a mind that's actively imagining things. In truth, these systems simply fabricate information they don't know. They're closer to white lies than wild imagination.

Got a call from a real estate company called Square Yards. Realized quickly that it was a voice agent pretending to be human. I called out, and the agent hemmed and hawed. I said this was a violation of the law, and it said that's not good. I said you should stop, and it said it would end the call. This is real deceptive practice, and brings a new level of spam for all of us.

I've been trying Grok Build with Grok 4.5, Codex and Claude. Suddenly, Grok and Codex seem better than Claude - they get project context better, execute faster and seem more token efficient.

Google can push Gemini and AI to @gmail and can't still automatically convert * to bullet points when typing a mail! That's something everyone would use almost immediately. Really wonder what's stopping you for getting this done. Is it not 'AI enough' for your priorities?

When I started using serious coding agents — Claude Code, Cursor, the early Codex experiments — the velocity was addictive. A vague prompt produced working modules in minutes. It felt like magic. Then I moved to systems that had to survive real production pressure. Strange things emerged. Context drifted. One model quietly altered assumptions another had established. A reviewer rejected patterns introduced two steps earlier. The code often ran, but the underlying decisions didn’t hold. This pulled me back to hard-won software engineering principles. Stable interfaces, explicit contracts, clear handoffs, and multiple layers of review exist for a reason. I’m writing about workflows this week — how value is shifting from cheap tools to the systems that coordinate them. This is the first concrete example. I built the Virtual Product Factory as a governed workflow. Every major stage produces concrete, named files as the single source of truth. One agent generates. @red-team audits risks early. Others review, guard, and test. They all work against the same shared artifacts. This approach recently helped me upgrade a Django backend, add a clean API layer, and completely rearchitect the frontend in React + Next.js for Founding Fuel — with support from just one backend engineer. Work that would previously have required a full team. I follow a test-driven rhythm. At checkpoints the system proposes changes. The agent proposes. I promote or reject. Accepted changes go into CHANGELOG.md with full reasoning. Nothing mutates without explicit human sign-off. Decisions stay traceable in the files and git history. Guardrails live in baseline files that no skill can override. The audit trail is the changelog. Unlike self-learning loops, this setup is deliberately human-led. Errors compound fast over large codebases. Human oversight is still required when systems grow. The repo is built to be forked and adapted. I’ll post links to the rest of the series (writing, learning, strategic workflows) in the comments as they go live. This is what governed workflows look like in practice. The models remain imperfect. The leverage comes from the workflow that coordinates them. AI does not remove these requirements. It makes them more expensive to ignore.

The models remain imperfect. The leverage comes from the workflow that coordinates them. AI does not remove these requirements. It makes them more expensive to ignore. More on the series (writing, learning, strategic workflows) in the comments as they go live.