AI, technology, education, and ideas about what comes next.
What happens to human expertise when AI makes intelligence cheap and scalable? This podcast episode explores Raw Intelligence, the Evaluator Economy, writing craft, AI model differences and the changing role of human judgment.
A satirical thought experiment: if humans, AI systems and robots ever become competing factions, can one clever human remain friends with everyone—and survive by backing whoever wins?
AI may not simply destroy jobs. It may remove the economic ladder through which ordinary people historically climbed—and force society to rethink work, ownership, distribution, purpose and freedom.
When AI stops doing the work and starts building the thinker, the goal shifts from immediate performance to lasting human capability.
When human intelligence and AI begin thinking together, the unit of intelligence may no longer be the individual mind—but the human–AI cognitive system.
Today’s AI is remarkably capable, but hallucination, sycophancy, reward hacking, weak scientific discovery, and limited memory reveal deeper design problems. What might the next generation need instead?
AI may not merely create artificial intelligence. It may create an economy in which human intelligence itself becomes scalable—through Mind Books, cognitive access, attribution and royalties.
AI is making intelligence, coordination and information cheaper. Could that eventually change how money, local economies and even government work?
AI may do more than replace individual tasks. By automating information processing, coordination and decision-making, it could compress the managerial hierarchy itself.
SEBI’s FY25–FY26 data reveal a much more complicated F&O picture than simply “most traders lose.” Capital, option strategy, prior profitability and experience all tell different stories.
What if computers evolved like nature builds complexity—by fixing the basic architecture and innovating above it? A proposal for standardized cassettes, cubes, and scalable computing infrastructure.
The hidden danger is not that AI cannot research. It is that it can research your question exactly as you framed it.
AI can be a Ferrari trapped on a one-lane road—but it can also be an extraordinary instrument in the hands of an ordinary user. The other AI bottleneck is the human ability to extract and direct its capability.
AI is making production cheaper. As generation becomes abundant, the scarce skill may shift to evaluation, judgement, context, cross-domain thinking and recognising what the machine missed.