OLDaily

Audiobooks or Reading? To Our Brains, It Doesn’t Matter
Jennifer Walter, Discover, 2019/08/30

According to this article, "new evidence suggests that, to our brains, reading and hearing a story might not be so different." I have serious doubts about this study (33 page PDF), and even more about the reporting. So much is left to interpretation. The study uses brain scan analyses, which we are told are "color-coded maps of the brain show the semantic similarities during listening and reading" based on volumetric pixels or voxels. But when we read some of us at least sound out the words in our mind (I know I do, some of the time). Why don't voxels (assuming they are a thing at all, which I doubt) just correspond to sounds? There's no reason to believe they carry any semantical import at all. Here's the phrenology, er., I mean 'pragmatic semantic atlas' (zip file from Box, .npz file (which is a file format by the Python numpy library that provides storage of array data using gzip compression)).

Web: [Direct Link] [This Post]

Hello World in Speech Recognition
Apoorv Nandan, Medium, 2019/08/30

This article is a bit complex (since it shows you how to write your own speech recognition system) but at the same time it's simple enough to allow us to grasp the essentials of one of the most intractable problems in learning: language. Let's take the simplest case possible: did a person say 'cat'? We solve this in two stages: we convert the audio to a feature matrix of digital data points called a 'spectrogram', then we feed this an input into a neural network that calculates the probability that we did indeed hear the word 'cat'. That's the gist of it. Everything else - how we prepare the audio, how we train the neural network, how we interpret the results - is detail. Messy messy detail.

Web: [Direct Link] [This Post]

Interpretations of Probability
Alan Hájek, Stanford Encyclopedia of Philosophy, 2019/08/30

This is an example I use quite a lot: the interpretation of probability. I use it to suggest that what counts as a fact depends very much on our point of view or perspective. Consider probability - consider, for example, the fact that there are 50-50 odds of heads or tails when we flip a coin. What do we mean when we say that? Is this a statement about the coin itself? Is it a statement about the universe as a whole (about, say, "all possible worlds")? Is it a statement about ourselves? As Bertrand Russell said, "Probability is the most important concept in modern science, especially as nobody has the slightest notion what it means." This article outlines six major interpretations of probability (I usually limit myself to three). None of these interpretations is wrong. None of them is right.

Web: [Direct Link] [This Post]

I tried to glitch the simulation and all I got was a bottle of pee
Tamlin Magee, The Outline, 2019/08/30

This is exactly what we do as educators. We find 'glitches in the simulation' - in other words, we point to 'attractors' that would otherwise be bits of random noise in the background of a person's existence. The person pays attention to that particular location, and finds something odd - a regularity, a set of identical cats, a green shopping bag containing a pair of Nikes propped up on a Burger King table. And between us we create a story that ties all these glitches together, a grand theory, if you will, about life and the universe. There's nothing in those things themselves that makes the theory true. But once we start accepting it as true, we begin to recognize it in everything. Even if it's just an attractor created by a random number generator. This, then, is the story of the Randonauts.

Web: [Direct Link] [This Post]

Education Technology and The Age of Surveillance Capitalism
Audrey Watters, Hack Education, 2019/08/30

This is a discussion of Shoshana Zuboff's book The Age of Surveillance Capitalism: The Fight for a Human Future at the New Frontier of Power. In it, Audrey Watters reaches back into B.F. Skinner's behaviourism quite a bit. Now as a philosophy student in the 1980s I learned (as did all others) that behaviourism is wrong. But I'm sensing a bit of a turn here in Watters's work: is behaviourism wrong because it doesn't work? Or is it wrong because it does work, and it's not ethical? Here's Watters: "increasingly, Zuboff contends, these companies are now using our data for much more than that: to shape and modify and predict our behavior." That sounds a lot like Zuboff is saying Skinner got it right. If we can be programmed even a little, what does that say about the relation between education, ethics and propaganda?

Web: [Direct Link] [This Post]