Six Links Worthy of Your Attention #669

Posted by

Is there one link, story, picture or thought that you saw online this week that you think somebody you know must see?

My friends: Alistair Croll (Solve for Interesting, Tilt the Windmill, Interesting Bits, HBS, chair of Strata, Startupfest, FWD50, and Scaletechconf; author of Lean Analytics and some other books), Hugh McGuire (Rebus Foundation, PressBooks, LibriVox) and I decided that every week the three of us are going to share one link for one another (for a total of six links) that each individual feels the other person “must see.”

Check out these six links that we’re recommending to one another:

  • A global aircraft-based wastewater genomic surveillance network for early warning of future pandemics – The Lancet“Tracking the spread of Covid, and other pathogens, is a tricky subject. Some countries don’t want to cooperate, and standards are messy. Pandemics spread fast in the modern age of air travel, too. Turns out that may present a solution (though the long title kinda spoiled the punchline): Measure aircraft wastewater. Many diseases can be detected in human waste — so why not use that as a canary in the disease coal mine? ‘Wastewater testing is the only current environmental surveillance technique that allows for understanding excreted human health indicators at the population level using a non-invasive sample.’ I love this idea, and since we discussed the math of Covid a lot over the past three years, I figured you would too.” (Alistair for Hugh).
  • Web LLM – Machine Learning Compilation. “Large Language Models are moving fast. One of the things they rely on is a GPU — turns out the things that power video games are really good at many parallel calculations, which makes them perfect for AI. But until recently, browsers couldn’t do math on GPUs. The experimental version of Chrome, called Canary, supports this now, which means you can use one of the open source language models and run a ChatGPT-like thing in your browser now. There are a couple of models like this (I’m sitting down with the folks who created one of them, Dolly 2.0, next week for a chat). It’s still a bit buggy, but this stuff is changing fast.” (Alistair for Mitch).
  • The idea that everything from spoons to stones is conscious is gaining academic credibility – Quartz. “What even is, like, consciousness, man?” (Hugh for Alistair).
  • Phonetic Alphabet. “Probably twice a month for the past 28 years, I’ve been on some phone call where someone is asking me to spell something or read out an alphanumeric code of some kind, and I panic once again for the 672nd time, and spell my name H-U-G-H, um, H for homunculus, U for, uh, uh, um undulate, G for, er, gu … ge…. go … gumdrops, and H for … [pause] … hibachi. By which time my face is flushed crimson and I am dripping with terrorsweats. So, as a public service announcement, I bring you this handy cheat sheet for how to sound cool when someone asks you to spell on the phone (Hotel-Uniform-Golf-Hotel). And I announce that I will dedicate the rest of my life to making learning the Nato phonetic alphabet off by heart part of the core curriculum for all high school students the world over.” (Hugh for Mitch).
  • Meditations On Moloch – Scott Alexander. “If you have been following the links here over the past while or listening to folks like Lex Fridman or Liv Boeree discuss anything related to artificial intelligence (and, for your own well-being, you really should) then you must have stumbled upon the concept of ‘Moloch’. Which, admittedly, was a new one for me, and is often sourced back to this article by Scott Alexander. The concept is… scary… obvious… and quite sad. I’ll do my best to transpose what it means: When humans race towards a point that has a payoff, but it also guarantees our destruction or is a lose/lose scenario. Another iteration is being convinced to play a game by something evil (coined “Moloch”) that is not good for humanity. Think about the nuclear arms race or how everyone knows that we have to completely re-invent the educational system, but we all just let it run, or even beauty filters. It’s explained much better by Max Tegmark in the following link. So… we know that we should stop doing something, but we simply can’t stop doing it. Ugh… the complexities of being human…” (Mitch for Alistair).
  • The Case for Halting AI Development – Max Tegmark – Lex Fridman Podcast. “As mentioned above and over the past many weeks, Lex Fridman has produced some of the most fascinating conversations about AI. Everything from why we should keep going to how it will destroy us to this conversation about why we should pump the brakes. I am still concerned about pumping the brakes or trying to stop AI (my reasoning is right here: Fear Of An AI Planet), but after listening to these conversations, learning much more about differing perspectives, and thinking about the issues critically, I am getting more and more concerned that we just might mess this up… and not get a second chance to get it right. If you think AI is just another type of technology, you are mistaken. Follow the links. Listen to these conversations. Allow your own perspectives to be challenged.” (Mitch for Hugh and video below).

Feel free to share these links and add your picks on TwitterFacebook, in the comments below or wherever you play.