What I Am Reading:

Reality is not made up of objects

Physicist Dennis Dieks makes a provocative argument here: quantum mechanics doesn't just complicate our picture of reality, it dissolves the idea of individual objects altogether. At the subatomic level, particles become indistinguishable and only exist in relation to one another. It genuinely changes how you think about what anything is.

Massive Gravity and the Dark Energy Enigma

What if gravity has mass? This piece lays out the theory of massive gravity as a genuine alternative to the cosmological constant. The idea is that even a tiny mass for the graviton could account for the universe's accelerating expansion without needing dark energy at all. It's a bold proposal and the author explains it clearly enough that it actually sticks.

The Anthropic Hive Mind

Steve Yegge spent years as a senior engineer at both Google and Amazon, so when he says something unusual is happening at Anthropic, it's worth paying attention. His argument isn't just about the models. It's about the organizational culture, the caliber of people, and the way the company thinks collectively. Interesting read regardless of where you stand on the AI race.

Meet the Startups Trying to Build Military-Specific AI

A wave of veteran-founded startups is now building AI trained exclusively on military and combat data, designed to work without cloud connectivity in denied environments. Companies like Smack Technologies just closed a $32M seed round on this thesis. The bet is that the Pentagon will never fully trust a general-purpose model for critical decisions, and that purpose-built systems are the only real path to adoption.

Defense Tech Startups Had Their Best Funding Year Ever in 2025

U.S. defense tech equity funding nearly tripled to $14.2 billion last year, with the number of firms investing in the sector up 41%. Ukraine validated a lot of what was previously theoretical about autonomous systems, and now the hard part begins. The bottleneck going forward isn't capital or even technology. It's production capacity.

Neutral Atom Quantum Computing: 2026's Big Leap

The case for neutral atoms as the winning qubit architecture keeps getting stronger. This piece covers Microsoft and Atom Computing's Magne system targeting 50 logical qubits by early 2027, while IBM is pushing to show quantum advantage before year end. We're getting to the point where the milestones are real and the timelines are short.

The First FDA-Approved Trial to Partially De-Age Humans

David Sinclair's Life Biosciences just got FDA clearance for the first human trial of partial epigenetic reprogramming. This is the same approach that reversed vision loss in primates. The framing I find most interesting is the economic one: with aging demographics globally, longevity isn't just a scientific goal anymore. The business case for solving it is enormous.


What I Am Listening To:

How to Design a More Meaningful Life

Dave Evans and Bill Burnett, the Stanford professors who wrote Designing Your Life, join Dr. Laurie Santos on The Happiness Lab to make the case that major life decisions shouldn't be planned, they should be prototyped. The idea is to treat your own life like a design problem: run small experiments, iterate, and let the data tell you what works. Practical in a way that most self-help conversations aren't.

Dario Amodei: We Are Near the End of the Exponential

Dwarkesh sat down with Anthropic's CEO in February and this one goes deep. Dario covers where RL scaling actually stands right now, how Anthropic thinks about the path to profitability, the economics of building at the frontier, US-China competition, and his honest view on AGI timelines. One of the better conversations on the state of the industry I've heard in a while.

Elon Musk: In 36 Months, the Cheapest Place to Put AI Will Be Space

Dwarkesh and John got extended time with Elon and covered a lot of ground: orbital data centers, why scaling power on Earth is becoming the real constraint, what it would actually take to manufacture humanoid robots at volume in the US, and how xAI fits into all of it. 

Future-Proof Your Brain from Dementia

A really thorough deep dive drawing on Dr. Tommy Wood's research, covering peptides and targeted lifestyle interventions aimed at protecting cognitive function over the long run. What I liked about this one is that the author isn't just summarizing studies, he's actually running the experiments on himself and reporting back honestly on what's working.

What I Am Watching:

Mammal Origins | NOVA | PBS

NOVA goes back 325 million years to trace where mammals actually came from, well before the dinosaurs. The story follows the ancient reptile-like creatures that gradually evolved warm blood, more complex brains, and the ability to nurse their young. It's one of those episodes that makes you appreciate how improbable and hard-won every feature of our biology really is.

The AI Doc: Or How I Became an Apocaloptimist

From the Oscar-winning director of Navalny and the team behind Everything Everywhere All at Once. A filmmaker who's about to become a father goes looking for honest answers about what AI means for the world his kid is inheriting. He talks to the believers and the skeptics and doesn't try to wrap it up neatly. Hitting theaters March 27 and probably the most timely documentary out this year.

Fukushima: A Nuclear Nightmare

HBO reconstructs the nine days following the 2011 Fukushima disaster with real forensic detail, putting you in the room with the engineers trying to prevent a full meltdown, the politicians who had to make impossible calls, and the people who just got caught in it. Worth watching right now given that nuclear is back in serious conversation as the answer to the AI energy problem.