annotate.mattgauger.com

I've been a fan of Prolog for a long time. There's some really good books, but this blog post will do as an introduciton.

The AlphaGo Zero -> AlphaZero generalization paper. A really interesting read.

Worth a quick skim, even if you have some background in ML.

It’s an incredible speed-up, and a perfect example of how AI can help astronomy.

Cool stuff.

What cognitive bias keeps us from realizing that unlocking intelligence amplification unlocks everything else?

At the beginning of each of these dinners, I would initiate a thought experiment:

What do we need to focus on today in order to create a world that we would love to live in by 2050?

With minor variations, I heard the same answers nearly every time — climate science, education, healthcare, AI, governance, and security. Not once, though, did any of the 150+ participants mention improving the brain itself. I was stunned by this. Why, I wondered.

And on understanding the difficulty of it:

There are difficulties ahead. Fixing the brain’s weak points in preparation for our increasingly complex future is a social challenge as much as a technical one.

I particularly like the phrase "The Imagination Economy."

Neuroprosthetics are on their way, possibly sooner than you think. And they won't necessarily require surgery:

… but he knows he’ll never sell a mass-market product that depends on drilling holes in people’s skulls. Instead, the algorithm will eventually connect to the brain through some variation of noninvasive interfaces being developed by scientists around the world, from tiny sensors that could be injected into the brain to genetically engineered neurons that can exchange data wirelessly with a hatlike receiver.

A look into the mind of Elon Musk. Check out Ashlee Vance's book Elon Musk: Inventing the Future for more background, if you haven't. There's also an extensive 4-part series on the Wait but Why blog.

If you've coded a business rule soup in methods for your OOP classes, and felt like it quickly became unmaintainable or something that you can no longer reason about, this might be an interesting read.

Some light reading for your Monday.

Another project attempting to build mesh networks with cheap/simple components and open source. This one uses ESP8226 boards and LoRa transceivers – so it might not have a lot of bandwidth but it should have some range and usefulness in the intended situation of disaster recovery. This project comes from SudoRoom makerspace and they have their own community mesh network built on OpenWRT, too: https://github.com/sudomesh/makenode

As made famous by Fred Brooks in The Mythical Man-Month. See also second system effect entry on the Jargon file. Another link that I've shared enough over the past few weeks to warrant an entry here.

What I'm currently working on to get my workflow for annotate.mattgauger.com down to less keystrokes.

by Richard Gabriel.

I had to share this at work again today, and I thought it was worth sharing here.

Being able to write a Bloom Filter in Erlang, or write multi-threaded C in your sleep is insufficient. None of that matters if no one wants to work with you.

In the realm of different programming paradigms, here's a way to do data flow programming on Node. It even has a visual output for the graph of components. (But NoFlo is not, itself, a visual programming tool / visual block language.) This reminds me a lot of the data flow programming I did on Hoplon, which is a data flow or "spreadsheet programming" library on ClojureScript.

Continuing on the theme of designing and operating internet-scale systems. The entire Google SRE book, which I read back in September, is worth a read.

by James Hamilton - Windows Live Services Platform

Currently reading this and thought I'd share.

While auto-administration is important, the most important factor is actually the service itself. Is the service efficient to automate? Is it what we refer to more generally as operations-friendly? Services that are operations-friendly require little human intervention, and both detect and recover from all but the most obscure failures without administrative intervention. This paper summarizes the best practices accumulated over many years in scaling some of the largest services at MSN and Windows Live.

For those that still use subscribe to RSS and want something that goes from their local emacs to their smartphone.

Some scifi for your day. By Max Tegmark, author if Life 3.0.

A DIY laptop built with an ARM chip. Putting together a laptop from a kit sounds really appealing to me right now.

I looked at this list of static site generators before deciding to use Org mode to generate this page.

This could prove very interesting. DARPA issued a Grand Challenge to use machine learning with software-defined radio a few years ago.