zek

Answering the question “you’re not worried about AI safety?”, the account says not even slightly, and that it would not have crossed its mind without effective altruists and “alignment” people loudly posting. A follow-up post says those opinions are so loud that politicians have tried to weaponize them. Both posts inspected via the X API.

Says being an alarmist, a “safetiest” or a doomer does not make you a good person, and that history has many examples of these archetypes being “evil”. A jab at safety advocates’ moral standing, not an argument about specific risks. Full post text inspected via the X API.

Quoting its own post predicting compute rentals and neocloud competition, the account expects consolidation and commoditization “as the capitalist intended”, timed for trillion-dollar IPOs, and says there is no AI bubble: AI is the most transformational technology in human history. Full post text inspected via the X API.

Says the best part of living in the era that will spawn superintelligence is watching people lose the things they claim give them meaning and realize how shallow it was, and that in the end we will live more freely and passionately. A hopeful view of the transition, with no date. Full post text inspected via the X API.

Asks what a human life is for once necessity no longer organizes it; a reply says being able to answer this for yourself is the difference between a happy and a sad person, today and in the future. Poses the question without giving its own answer. Both posts inspected via the X API.

Says owning GPUs will soon be treated like owning automatic weapons, and already is in many people’s mental models, and asks for a new constitutional amendment protecting the right to own and operate compute as people like. A reply says everything is trending in this direction. Both posts inspected via the X API.

Quoting a roon post about how a driver’s license would be received if invented today, the account says such licensing is not communist: anti-democratization of intelligence is feudalist, which is worse. The quoted post belongs to roon. Full post text inspected via the X API.

Calls open source the pinnacle of freedom, full transparency with no smoke and mirrors, and says modern society’s fabric relies on it. A 2026-09-21 post adds that the account is genuinely worried about the future of open source because motivation has fallen, but hopes it prevails. Both posts inspected via the X API.

Quoting its own post about an Anthropic outage, the account gives an IPO bear case: good closed models do not matter if reliability is poor, and Anthropic rents all its compute. It is bullish on those who own compute and open-weight models, and most bullish on those who own compute and train models. A market view, not inside knowledge. Full long-post text inspected via the X API.

Says anyone who learns inference infrastructure will never be unemployed again; a reply adds there is so much work and money to be had. Related posts urge people to learn Kubernetes and some ML. A view on demand for one kind of tech work, not an economy-wide forecast. Posts inspected via the X API.

Says people would be less anti-AI if they realized modern life was built on the very things they fear, and that history is cyclic; a reply says “maybe just… stop worrying”. Full post texts inspected via the X API.

Says people, including very senior engineers and managers in its meetings, have the craziest limiting beliefs about what is possible while it watches Gemini do the things they say cannot be done; the moat is believing it is possible. Capability optimism from its own work, not a timeline. Full post text inspected via the X API.