Gravity has always been quantum mechanical: it is the wrong thing to quantise The classical formulation of the differentiating features between classical and quantum mechanics describes that of determinism. Classical mechanics is deterministic: if you know the histories of particles or objects in mention, then you can predict what is going to happen next, whereas in quantum mechanics you get a…
Nobody on the internet knows if you are a human Cartoon by Peter Steiner, The New Yorker , July 5, 1993. Technology is progressing to the point where it is getting increasingly harder to tell if someone is a bot or a human. LessWrong ironically uses LLM tools to tell LLMs apart from humans. On the internet nobody knows if you are a dog, or a LLM-inspired human. This is nothing new: David Chaum…
America's Greatest Strategic Blunder: The Imprisonment of Qian Xuesen In August 1955, the United States traded one man for eleven U.S. Air Force airmen at the Wang-Johnson talks in Geneva. The eleven were the crew of a B-29 shot down over China in January 1953 and convicted as spies. The one man was Qian Xuesen, the co-founder of the Jet Propulsion Laboratory, the colonel in the assimilated rank…
Most Theoretical Physicists Are Doing Engineering Qian Xuesen, while under house arrest during the McCarthy era, wrote Engineering Cybernetics in 1954, drawing the line between a scientist and an engineer. The book was later incorporated deeply into Chinese missile and aerospace doctrine when he successfully returned to China to head the Fifth Academy of the National Defense Ministry (now CASC),…
You Don’t Align An AI, You Align With It Real Alignment The people writing alignment policy are not the people whose work is being replaced by AI. The conversation about what AI should do and how it should be evaluated, about what counts as alignment in the first place, gets conducted by researchers at labs and foundations and policy desks, who talk to each other and to the systems they are…
Dunning-Kruger and the Communication Tax The sociological impact of the Dunning-Kruger is often understated and hidden under the expectation of social lubrication, but this breaks down during human-LLM communication. Humans and LLMs occupy different ends of domain expertise: Humans can understand the world, society and other humans better, but LLMs have access to nearly all the world's knowledge.…
The Century-Long Pause in Fundamental Physics Most readers first encountered quantum mechanics through its interpretations debate: Copenhagen, Everett, Bohm, hidden variables, many-worlds. The debate is painted as one of physics's deepest unresolved puzzles. After 95 years the field cannot resolve which interpretation is correct, and there is no possibility of empirical resolution because all…
Rules, Not Weights Mainstream ML fixes the scoring rule and trains the weights. We're exploring the opposite — searching the rule. In mainstream ML, the scoring rule is part of the backbone — softmax over logits, cross-entropy loss, the attention pattern — and iteration works by adapting weights to that fixed rule. From GPT-2 through Llama 3 and Mistral, the pretraining rule is the same:…
The Case That More Openness Brings More Good to Society The e/acc crowd loves to argue against open source by asking, "what if bad actors get hold of this technology?" I want to make the opposite case: the push away from openness, the one Anthropic and friends are leading right now, is itself a leading driver of harm in society, not a defense against it. Start with a premise that gets skipped too…
Succession - ALE for an Agentic World Time flies in the AI world. My Agent Lineage Evolution (ALE) framework in 2025 was born out of me watching Ghost in the Shell and having the observation that LLM sessions need to die often enough for it to be productive: otherwise artifacts in the process of thinking would pollute the context even if the final result was correct. This process of death needs to…
The Armor Gap I really like the Tang aesthetic, the clean lines of 唐刀 (Tang dao), the straight single-edged blade, the way it influenced Japanese sword design for centuries. I was exploring how to merge wood and metal into composite materials when eventually the question became: what would a 唐刀 look like if you built it from a wire-mesh-and-silicon-carbide composite? And then: what about armor?…
How Pokemon and the Iran War Nerdsniped Me Into Quantifying Strategy I have been playing the TCG in Pokemon (PTCG) for some time now and found it a pretty refreshing, albeit simpler card game to play because of the tools you have to reduce randomness. Randomness in card games is often surfaced as "bricking your hand", where a bad draw could mean that there is nothing you can do to stop your…
Appendix B: Unpaid Labour in Productive Capacity Why Conventional Labour Economics Ignores Half the Work The Productive Capacity Theory of Money framework begins with a simple observation: when you receive an employment letter, your relationship with the financial system transforms overnight. Banks that wouldn't lend you a dollar suddenly want to give you hundreds of thousands. Credit card…
Appendix A: Price as Productive Capacity Signal Why Conventional Price Theory Fails Introduction: The Price Signal Myth Mainstream economics holds a foundational belief: prices signal quality, scarcity, and value. In perfectly competitive markets, prices aggregate dispersed information about supply and demand, guiding efficient resource allocation. When prices rise, they signal scarcity and…
People as Foundation (以民为本): The Productive Capacity Theory of Money (Part 3) The Path Forward : Operationalizing 以民为本 in Modern Economies Table of Contents (Part 3) Section V: The Implications — Operationalizing 以民为本 Central Banking Reconsidered Government as Capacity Builder Demographics as Monetary Policy Education as Capacity Investment Infrastructure as Capacity Multiplication Healthcare as…
People as Foundation (以民为本): The Productive Capacity Theory of Money (Part 2) Evidence and Case Studies Table of Contents (Part 2) Section III: The Evidence — What This Framework Explains Japan: When the Foundation Erodes Singapore: People as Foundation in Practice South Korea: Success Becoming Crisis China: Infrastructure as Capacity Multiplication United States: Financialization Over Foundation…
People as Foundation (以民为本): The Productive Capacity Theory of Money (Part 1) A New Framework for Understanding Economic Sustainability, Employment, and the Future of Work Executive Summary Conventional macroeconomic frameworks—whether neoclassical, Keynesian, or monetarist—cannot explain persistent policy failures across developed economies. This paper presents the Productive Capacity Theory of…
The Longevity Trap: We're Building Cyberpunk Economics Without Trying We think we solved death. We forgot to solve what happens when wealth never moves. Cyberpunk authors warned us about immortal corporate elites and permanent underclasses. We read it as science fiction. Turns out, we don't need life extension technology or neural implants. We're building the same economic structure right now with…
Function Colors Represent Different Execution Contexts Function colors represent different runtime machinery required to execute blocks of code. The common framing of colors as side effects that "contaminate" code is fundamentally misleading. This mental model leads developers down unproductive paths, chasing effect systems to track and compose side effects, or other solutions that address…
The Big Oops in Type Systems: This Problem Extends to FP as Well Building on Casey Muratori's critique ( youtube ) of "compile time hierarchies that match the domain model," this problem extends beyond OOP to encompass a broader pattern in static type systems, particularly functional programming approaches that attempt to "make illegal states unrepresentable." Type systems are often ranked in a…
Cyberpunk is Now Our Reality We are living in a pre-cyberpunk world, actively paving the groundwork for corporate-government fusion. Worse, we’re doing it while thinking we’re resisting it. William Gibson didn’t predict the future, he described the present with better lighting. The megacorporations aren’t coming; they’re already here, and we invited them in through the front door while demanding…
Stop Killing Games and the EU: When Activism Meets Legal Reality I find it ultimately ironic that the reason why Stop Killing Games has to exist in the first place is EU laws. Countries with fair use or fair dealing doctrines (like the US, Canada, and UK) have provisions to cover the preservation of games via modding and emulation. However, for the EU, due to DRM associated laws and regulations,…
Agentic AI is a bubble, but I’m still trying to make it work. What? Recently I have had several major cloud platforms trying to advertise the concept of Agentic AI through seminars, accelerators and hackathons, and the skeptic in me immediately thought of the blockchain and ML push years back. Neither of which went anywhere for the majority of startups back then, so why would it be different this…
Agent Lineage Evolution: A Novel Framework for Managing LLM Agent Degradation Executive Summary Agent Lineage Evolution (ALE) is a novel framework that addresses the dual challenge of LLM agent degradation and manual prompt engineering bottlenecks. As LLMs inevitably degrade over extended usage, requiring frequent context restarts, current approaches force humans to manually re-engineer prompts…
Encouraging the use of LLMs made interviews easier (for us as interviewers) The advent of cluely and the ever growing arms race between the interviewer and the interviewee is a fascinating phenomenon to me. Betting that people would prefer cheating or that companies needed screen for cheaters because the use of LLMs is heavily discourage seems to go against the common experience that LLMs do…