I think about AI’s business reality through technical depth, quality standards, user trust, and whether companies can turn AI into real revenue.
On 2025-08-22, I argued from a builder and investor lens that many exciting AI startups and demos are quickly assembled API wrappers marketed as innovation, and that wrappers rarely scale. I see defensibility in engineering-led startups that understand models and frameworks under the hood and can fine-tune, optimize, and push them beyond their limits. Wrappers may impress in the short term, but I expect deep builders to shape the future. I left open whether wrappers can survive or real innovation will win.
On 2025-10-23, I observed that people with a strong internal bar for quality often do not find GenAI magical or transformative: they still have to put in the same effort to meet their standards. In my view, GenAI’s magic is more apparent to people who never had a clear definition of quality; to them it feels like a breakthrough, while to others it is a faster draft.
On 2026-09-01, I asked whether people were losing more time to what I called “basic lies.” I did not identify the errors or provide examples, so the question remains open.
On 2026-09-30, I described an AI crash as investors realizing AI is not making real money, causing tech stocks to plummet and wiping out trillions in market value overnight. I gave no specific amount or criteria for what counts as AI making real money.
Related: running an early stage startup, investing in startups from india, upskilling workers and social impact