Shiju Thomas

Once you hit about a 20-point IQ gap, communication starts to completely break down. It's not that the lower IQ person is "stupid" (although that can often be the case) or the higher one is arrogant, it's that you're literally operating on different systems. A 20 point difference (roughly 1.3 standard deviations) means: Vocabulary and abstraction levels diverge sharply. What feels like crystal clear logic to one side sounds like vague, pretentious word salad to the other. Jokes land flat. Metaphors get taken literally. Complex cause and effect chains get simplified into "this good, that bad." Different time horizons and pattern recognition. One person thinks in months or years and sees systems, the other is locked into days or immediate rewards. Trying to explain second order effects feels like speaking another language. Also, processing speed and working memory gaps. The higher IQ person is already three steps ahead, getting impatient. The lower IQ person feels talked down to or overwhelmed. Both walk away frustrated. Both have wasted each others time.

I had stopped being able to climb the stairs. Not pain, exactly. A kind of polite refusal.

The book I was not writing was more useful to me than the book would be.

Google Search Console picked up all 1,032 pages within minutes of the sitemap resubmission. If your older Lovable site is stuck on "Discovered, currently not indexed," the host is likely the wall.

The fix for us was moving to Netlify, which took one config file and a DNS change. Netlify serves static files first and only falls back to the SPA for routes with no prerendered match.

We built a 1,032 page career site on Lovable and Google indexed zero of them for months, because every URL served the same empty homepage shell to crawlers.

Every broken promise is an open tab in the browser of your mind. You don't notice the drain until you close a few.

@KrzysztofStaron This is pretty neat, lots of applications apart from fnding the best dev

There are 47 milion programmers in the world, And with the same math hedge funds use to find undervalued assets it's possible to find the best ones. Data is scattered across personal blogs, stack overflow, github, reddit, TikTok, database leaks, employment contracts... it's about 150 TB of unstructured nodes of data. But after 1000's of processing steps required, and building a virtual satelity to spy on everyone, finding the perfect hire is as simple as writing a prompt. ( or SQL query ) ( Prototype in thread )

@TimoBuilds_ Absolutely - always reminds me of Blackberry, their customer feedback was for better keyboards, which got shipped but didn't save them when the keyboard went away completely / the opposite was Slack, the messaging platform for work which no one asked for, coming out of a game

@the_saint This is why I think “feature request” is often the wrong label. Most of the time it’s a user trying to describe a pain with the vocabulary they have. The founder’s job is to track the pattern behind the request, not blindly ship the proposed solution.

Customers do not actually want features even though they will happily give you feature requests if you ask. What they want is their problems solved, and feature requests are just their amateur attempt at solution design. Every feature request is a symptom pointing at an underlying problem, and your job is to diagnose the real issue. Sometimes the right solution is new code, sometimes it is better documentation, and sometimes it is doing nothing at all.

Series A companies typically die from lack of product-market fit because they never found something customers genuinely wanted, but Series B companies die from lack of go-to-market fit because they found a product people wanted without finding a repeatable way to reach those people. Having a product customers love is necessary but nowhere near sufficient for building a large company. Product-market fit is the lock; go-to-market fit is the key that actually opens the door.

Twenty browser tabs open at 2 am, each running a different AI tool. I tell myself I'll close the laptop after the next command. I do not.

Keep OpenRouter for secondary models. Route cheap tasks to Gemini Flash. But the orchestrating model should talk to Anthropic directly.

Fix: direct Anthropic API key in OpenClaw config. Set api to anthropic-messages. Prompt caching, native tools, full context window. Two-minute change.

Running Claude in OpenClaw through OpenRouter? You're paying full price for a stripped-down model.