Monthly Archives: June 2019

So Shelley and I are talking …

About age-related decline. She had a depressing get together with friends and they were all sharing stories of aging, none particularly uplifting. And last night there’s a thing on the radio about diabetes, and I understand that, genetically speaking, it’s probably something I’m going to be dealing with in a few years, and … on and on.

And lo and behold, next morning I’m reading this, and thinking that he’s onto something. Stuff shows up when you’re ready to pay attention, it seems.


The most thoughtful post I’ve read recently on these two apparently related (problem + solution) ideas is Vi Hart’s.

Her conclusion: AI is going to displace jobs, but it’s not what you think. UBI is a good idea, but not as a solution to AI. A better solution is based around fairly valuing people for the data they generate, which is fundamental to the successful operation of AI. Lots of good references to Lanier, et al.

Definitely worth the read if you’re interested in this. Andrew Yang, do you have a good response?

Facebook and news

I guess I don’t understand the hue and cry over Facebook. The company is clear that it doesn’t, and doesn’t want to, act as an editor for the variety of things that get posted, and it’s clear that given the volume of stuff they really can’t do a good job at it. Meanwhile, there are sources of information that do have editorial standards. As a consumer of information, I get to decide: do I want to get my information from sources I trust to vet and present it truthfully, or do I want to get it from anyone?

As a matter of fact, I do both. I read things on Reddit, but I approach them with much more skepticism than I do with my New York Times subscription or Richard Stallman’s blog. By the same token, I subscribe to Taran’s Free Jazz Hour, because of his curatorial excellence.

So, is the problem just that people have an unreasonable expectation of Facebook?