Notes of Nagesh

  World Model: Mind over matter AI companies want to build machines that understand the physical world. If they succeed, it will be the most consequential leap in the technology since the transformer. CHUPPALA NAGESH BHUSHAN   |  Jun 12th 2026 | HYDERABAD I MAGINE asking someone for directions in an unfamiliar city. If they know the place, they can improvise, reroute around a closed street and adapt on the fly. If they are merely repeating a memorised script, a single detour leaves them helpless. Today's AI systems, for all their dazzling fluency, are closer to the second kind of navigator. A new generation of research aims to produce the first. The idea goes by the name of a world model . In its simplest form, it is an internal simulation of reality: a mental map of how objects move, how causes produce effects and how actions ripple through an environment. Humans and animals build such models continuously; they are why you do not need to stub your toe t...

From bags of chips to the Windows of war How a scrappy Pentagon AI project quietly became the backbone of American military power CHUPPALA NAGESH BHUSHAN     Jun 11th 2026   Drew Kukor arrived in Afghanistan in October 2001 lugging a heavy laptop two months after the September 11th attacks. As a Marine intelligence officer, he found himself operating in a near-total information vacuum: patchy data, dysfunctional analytic tools, and a war effort that was recording intelligence in Microsoft Word. The experience left a mark. Two decades later, the project he would champion—a Pentagon AI initiative called Maven—has transformed how America wages war. Project Maven began with a humble premise: the United States military was drowning in drone footage it could not watch. Thousands of hours of video poured in from unmanned aerial vehicles patrolling Iraq and Afghanistan, but the analysts tasked with reviewing it were overwhelmed. Screens flickered unobserved. Acti...

   The Musk Doctrine How a man who grew up under apartheid came to sell sovereignty-as-a-service to nation-states, rewire American defence and, in the process, invent something that looks disconcertingly like a new ism CHUPPALA NAGESH BHUSHAN   Jun 11th 2026 W hen historians of the early twentieth century sought to make sense of the social upheaval wrought by mass production, they did not merely catalogue Henry Ford’s eccentricities. They coined a term—Fordism—and used his factories and habits as a prism through which to read an entire civilisational shift. A similar intellectual exercise now presents itself with far greater urgency. Elon Musk is not simply a flamboyant billionaire with a weakness for social-media provocation. He is, argue Quinn Slobodian and Ben Tarnoff in their book  “Muskism: A Guide for the Perplexed,”  the load-bearing column of a new and distinctly unsettling political-economic order. The one-line definition the authors offer is decep...

  The invisible architecture of potential Why schools and clinics alone cannot build the human capital that poor countries need Chuppala Nagesh Bhushan ⅔ Share of income gap between rich and poor nations explained by human capital 70% Workers in developing countries stuck in low-learning jobs 43% Potential GDP gain for India from removing barriers to women's employment Human capital—the health, knowledge, and skills people accumulate over their lifetimes—is not built through formal sectors like education and health alone; it is built through daily interactions and decisions within homes, neighbourhoods, and workplaces. These settings collectively shape an individual's trajectory because human capital formation is a cumulative process where skills acquired in one stage facilitate learning in the next. In the twentieth century, the average height of a Western European adult rose by roughly one centimetre per decade—a sturdy proxy for improvements in population health. In several ...

  The Machine That Builds Itself More than 80% of Anthropic's software is built by its own AI. The recursive era has quietly begun. AI is now writing the code that makes AI smarter. The question is whether humans can stay in the loop. Welcome to "Recursive self-improvement" Chuppala Nagesh Bhushan · >80% of Anthropic production code written by Claude, as of May 2026 8× more code merged per engineer per day vs 2024 75–80% accuracy on SWE-bench Verified, frontier models in 2026 For decades the prospect of an "intelligence explosion" has hovered at the edge of serious technological discussion. An artificial-intelligence system capable of redesigning and improving itself, iteratively spawning ever more capable successors, could one day outstrip human comprehension and control. What once seemed a distant theoretical concern — popularised by the mathematician I.J. Good in the 1960s — has moved closer to the present. "As of May 2026, more than 80% of the code m...