Turns out Buffett was right again
The only stock market-level indicator he considers of any value is Stock Market Cap / GDP
Surprisingly, that had never been verified by academics, until now.
papers.ssrn.com/sol3/papers.cf…
Mackenzie Morehead
3,302 posts
Mackenzie Morehead
@mackenziejem
Joined November 2014
- New Post on Automated Science After years of studying the space, we at @CompoundVC believe the pieces are finally in place to autonomously conduct novel, high efficiency/throughput science In fact, we'll take the under on the Nobel Turing Challenge, which sets 2050 as the goal
- New post on biomanufacturing Despite the sector being so out of favor, we @CompoundVC see it emerging from disillusionment Synbio allows for novel science-driven consumer, cheaper biologic drugs, and dozens of Nobel-worthy tech await to disrupt mass produced goods
- You are really not taking seriously enough our increasing ability to replace experiments entirely with computation for all scales and for everything from bio to materials sciences to physics Selection of papers in just the last couple months:
- I just published a new piece on next-gen compute moonshots harnessing new physical systems - overview of the next decade in digital compute - AI ASICs (photonics, stochastic/thermodynamic, compute in-memory) - Niche HPC (quantum) - Edge intelligence (C-I-M, compute in-sensor,
- New Post: Cryptography's in a 0 to 1 moment. A generational transition from special- to general-purpose cryptography Probe a black box and get an answer cryptographically guaranteed to be correct, while keeping all the data & computation fully private We detail the applications
- New post on the near frontier of Neural Net Potentials and the implication for company building and industry competitive dynamics These models are smaller than GPT1. Excited to see them scale Thoughts on: * role in computational stack * materials science vs drug discovery *
Thesis: deep learning and ground truth physics are increasingly converging, and that gap will be closed primarily by better physics modeling, not AI In other words, the future is training models on data generated from accurate + computationally efficient approx. of the physics - Just finished Capital Wars: The Rise of Global Liquidity by @crossbordercap. One of the best books on finance / economics / investing I’ve ever read. Absolutely essential reading for anyone operating in those professions in the 21st century. Well done sir.
- Thesis: deep learning and ground truth physics are increasingly converging, and that gap will be closed primarily by better physics modeling, not AI In other words, the future is training models on data generated from accurate + computationally efficient approx. of the physics
- New post on biochemical wearables, ingestibles, and implantables For the first time these devices enable us to measure complex disease states and then respond in real-time From hormones to microbiome to BCIs to diabetes These devices move us from FitBits measuring simple
- I published Part I of our thesis on building a company in biosecurity: Why It Matters Pandemics have killed 230-400M ppl w/ COVID killing 27M in 4yr and costing $10Ts Add in infectious disease (incl cold, flu) and airborne viruses are an all-time driver of suffering and econ drag
- Replying to @mackenziejemPsych one more: @hussmanjp performed the same task across a variety of metrics for the US specifically, finding that a slight tweak to the straight up Market Cap / GDP used above provides even better results for the US:
- On April 24th in SF, Compound is hosting a Research Day on the scientific frontiers of wearables, implantables, ingestibles We have an incredible lineup of speakers If you'd to attend the event or know anyone who would, pls respond to the form below
- Replying to @mackenziejemWe've seen exciting proofs of concept but none that have truly pushed the bounds of science In Part II, we'll talk through the types of business models that can be employed and what commercialization approaches we’re most excited about
