• Hyperlegibility by Packy McCormick

    • Hyperlegibility emerges with game theoretical certainty from each of our desire to win whatever game it is you’re playing. Certainly, it’s a consequence of playing The Great Online Game. In order for the right people and projects to find you, you must make yourself legible to them. To stand out in a sea of people making themselves legible, you must make yourself Hyperlegible: so easy to read and understand you rise to the top.

    • Once you become aware of Hyperlegibility, you see it everywhere

  • YOUTH And what happens when it’s gone by Jared Young

    • To do a thing young is a shortcut to greatness. Everything you do right is evidence of your great potential, and all your mistakes are the mistakes of inexperience, which will be corrected, surely, by the passing of time. To write a poor sentence at the age of nineteen is expected (and quickly forgiven). To write a poor sentence at the age of forty-five—shouldn’t you know better by now? In this manner, being a not-young writer is a lose-lose proposition, because this principle applies to brilliant sentences, too. In youth, they are a marvel; later, they’re table stakes. Couples, which John Updike wrote when he was thirty-six, is considered one of his great accomplishments. But Villages, which he wrote at the age of seventy-two, is just as good. But of course it’s good—it was written by the guy who wrote Couples!

  • Don’t compete by Paras Chopra

    • You see, ultimately the society stands to benefit from this competition between people. Highlight and celebrate the lottery winner to fool enough people into competing with each other to make everyone life better.

    • Rather, what this implies is something very simple – don’t confuse what gets social approval with what’s right for you. Social approval exists to attract participants in a game that ultimately benefits the collective at the expense of an individual.

  • Kill your Feeds - Stop letting algorithms dictate how you think

    • With the power to shape what we see comes the power to shape what we believe. Whether through deliberate manipulation or the slow creep of algorithmic recommendations, engagement is fueled by outrage, and outrage breeds extremism. The result is a feedback loop that isolates users, reinforces beliefs, and deprioritises opposing viewpoints.

    • The internet should serve you, not the other way around. Take back control. Kill your feeds before they kill your ability to think independently.

  • A visual introduction to Gaussian Belief Propagation

    • In this article, we present a visual introduction to Gaussian Belief Propagation (GBP), a probabilistic inference algorithm that operates by passing messages between the nodes of arbitrarily structured factor graphs. A special case of loopy belief propagation, GBP updates rely only on local information and will converge independently of the message schedule. Our key argument is that, given recent trends in computing hardware, GBP has the right computational properties to act as a scalable distributed probabilistic inference framework for future machine learning systems.

    • Well written article

  • Mathematical Foundations of Reinforcement Learning

    • This book aims to provide a mathematical but friendly introduction to the fundamental concepts, basic problems, and classic algorithms in reinforcement learning.

  • Tracing the thoughts of a large language model (Anthropic)

    • Today, we’re sharing two new papers that represent progress on the development of the “microscope”, and the application of it to see new “AI biology”. In the first paper, we extend our prior work locating interpretable concepts (“features”) inside a model to link those concepts together into computational “circuits”, revealing parts of the pathway that transforms the words that go into Claude into the words that come out. In the second, we look inside Claude 3.5 Haiku, performing deep studies of simple tasks representative of ten crucial model behaviors, including the three described above.

  • OpenWorm: Building the first digital life form. Open source.

    • Larson is a cofounder of OpenWorm, an open source software effort that has been trying, since 2011, to build a computer simulation of a microscopic nematode called Caenorhabditis elegans. His goal is nothing less than a digital twin of the real worm, accurate down to the molecule. If OpenWorm can manage this, it would be the first virtual animal—and an embodiment of all our knowledge not only about C. elegans, which is one of the most-studied animals in science, but about how brains interact with the world to produce behavior: the “holy grail,” as OpenWorm puts it, of systems biology.

  • Large Bio-Mechanical Space Structures (DARPA)

    • Given recent advances in metabolic engineering for rapid growth, extremophiles with novel properties, biological self-assembly properties of tunable materials, and emergent mechanical design principles of biological systems, DARPA is interested in exploring the feasibility of “growing” biological structures of unprecedented size in microgravity. Rapid, controlled, directional growth to create very large (500+ meter length) useful space structures would disrupt the current state-of-the-art and position biology as a complimentary component of the in-space assembly infrastructure. Some examples of structures that could be biologically manufactured and assembled, but that may be infeasible to produce traditionally, include tethers for a space elevator, grid-nets for orbital debris remediation, kilometer-scale interferometers for radio science, new self-assembled wings of a commercial space station for hosting additional payloads, or on-demand production of patch materials to adhere and repair micrometeorite damage.