Saturday, May 2, 2020

The Math(s) Fix: An Education Blueprint for the AI Age by Conrad Wolfram


Rating: WARTY!

From an advance review copy for which I thank the publisher.

At one point in this book, the author writes, "When I started this journey, I thought there would be a huge amount of straight hostility. So far, I've found confusion predominates instead." Having read a substantial portion of it before giving up on it, I can only agree.

The book is written by the younger brother of the creator of the Mathematica software, and given that this very software is mentioned more than thirty times throughout the text, I had to wonder if this really is nothing more than an extended sales pitch for said software. The truth is that I honestly cannot say because despite the book being billed as "a groundbreaking book that exposes why math education is in crisis worldwide and how the only fix is a fundamentally new mainstream subject" I could not for the life of me, despite several searches throughout the book, discover what it is that the author proposes to replace traditional math teaching with.

That said, I must confess that I gave up on it about 25% of the way in. The book really dragged. Instead of launching into the new ideas from the outset, the author requires that we spend fully a quarter of the book listening to him waffling on about the problem without really telling us anything. I agree with him that the math we teach these days has little to do with most people's real-world experience of or need for it. The simple solution to that is to teach less of it and more of what people do need!

The language of this book is a bit high level, too. I wasn't sure who the author's intended audience was supposed to be, but given the college-level language he uses, it's definitely not the stereotypical 'man (or woman) in the street'. I didn't have too much trouble understanding most of it, but the writing was very dense, and quite academic in tone. I listened to it (read by my iPhone's Voice Over software) on the commute to and from work each day, and on the morning I decided to give up on it, and the reason I quit was because I realized that I had not understood a single word he'd written in some twenty-five minutes of driving.

This was not because I was too focused on traffic. The streets are largely devoid of traffic when I drive in to work, and I typically have no problem driving safely and hearing what my book or novel of choice is all about as I drive. That morning was a huge fail in this regard, and it's solely because of the high-falutin' language he used.

I read scores of books of all types, and have college-level education, and while it was not wholly impenetrable, this book was far too dense for my taste. He could have eased this quite readily by employing more everyday language, but his attitude seemed to be "why use 'used' when you can write 'utilized'"?! I can't take anyone seriously who regularly writes 'utilized'. For a book that claims to be clearing the cobwebs out of mathematics, perhaps his first step should have been to clear the cobwebs out of his writing, and write at a level that's easy for your average reader to grasp? Just a thought!

Just so you know it's not only me, I pasted the first 600 or so words from the first chapter into an online readability app, and these were the results:

  • Flesch Reading Ease score: 39.1 (difficult to read)
  • Gunning Fog: 17.2: (difficult to read)
  • Flesch-Kincaid Grade Level: 14.1 (College)
  • The Coleman-Liau Index: 13 (College)
  • The SMOG Index: 12.4: (Twelfth Grade)
  • Automated Readability Index: 15.6 (College graduate)
  • Linsear Write Formula : 17.3 (College Graduate and above)
So: not aimed at Jo Average! But it wasn't just the level of the language, it was the jargon employed. The word, 'computational' for example appears over 400 times. Here are a few examples, and no, I did not bookmark these at the time (driving!) I just went to random places in the book, and swiped a page or two in one direction or another, and sure enough there was a phrase right there. It's not hard to find them:

"Nor do they provide an appropriate structure for so doing, though in some cases they're complementary outcomes lists and can usefully coexist with outcomes for core computation."

"...that's nullifying the point of having a machine do it instead..."

"...indeed, that the rationale is not orchestrated for practical application distinguishes the discipline..."

"One of the drivers for this is the aforementioned problem of traditional outcomes listings being per maths tool, where our outcomes map instead reflects a distillation of substructure..."

"...not pre-abstracted calculation problem segments..."

"...with respect to a a (sic) core computational curriculum change..."

If only some of this had been rendered into more everyday language it would have improved readability immensely. But this was not the worst problem for me.

The real problem I had was that I really wanted to know what his alternative was, and beyond a vague idea that it seems to involve using computer software, I could glean no idea from the opening portion of the book, and nothing from skimming through and doing some reading in later sections to see if it's explained anywhere at all. I confess it's entirely possible, not having read the whole thing, that I could well have missed it, but I could not for the life of me find anywhere where the author says, 'this is what I propose' or words to that effect and lays out a summary of the new plan. The fact that this book has no contents page did not help in my forlorn quest to get to the 'core computation' (to use a phrase of the author's) and find out what he would like to see as the future of math education. To me that was a serious failing.

Given how tedious it was to read this, and how the author himself seemed curiously loathe to share his plan with the reader, I can't in good faith commend this as a worthy read. The problem seemed to be that he was preaching to the choir for the first quarter of the book. If the language had been simplified a bit, and he'd ditched that first 25% and launched right into it, assuming his readers were interested not in the sorry history of math education, but in discovering what his new proposal was, he would have made a better impression on me. But if his only plan is to sell the Mathematica software to every student at eighty bucks a year, then this seems a little self-serving to me. Maybe he had some other plan; I can't say because I couldn't find out what his plan was!