Jan

25

Two Professors at the University of Washington are developing a course based on BS detection. The syllabus has some useful reading in the links.

Mr. Isomorphisms writes: 

I'm skeptical that professors can teach students how to recognize bullshit. This used to go under the name "critical thinking", which is what liberal education has claimed to teach for a long time.

Education levels are now higher than ever before; is there less bullshit or better critical thinking about it, than in decades past when education levels were lower? Why not?

And why is "big data", among all other bullshit, such a powerful buzzword today?

The Decline & Fall of IBM by R X Cringely has partially answered my questions about why a former economic keystone has abandoned all reason and now churns out cognitive "analytics", which I'm sure we can all agree does not make sense or exist.

"IBM tells the customer what to do, not the other way around" is Cringely's description of the dancing elephant. A partial answer to the obvious follow-up questions is that (as Herb Simon noted 50 years ago) the dynamics of large organizations/ teams are what drive output, not "market" forces as normally construed. Your promotion does not depend on whether the customer likes your work, but whether your boss's boss does. "Cognitive" and "behavioural" are good signals of bullshit.

It's fairly clear to me why Facebook and Google are funding the big-data-analytics-machine-learning movements: They're monopolists with large ad revenues and special share classes; they don't get punished by their shareholders for weird hires (eg, Ray Kurzweil).

There is a compelling case for the self-driving lorry, but really Google X is doing whatever it wants; that one or more of the blue-sky projects could theoretically benefit shareholders at some point is not the reason any of them are funded.

IBM, FB, and GOOG then hire academics who otherwise have no useful skills out of their universities, thus driving demand for machine-learning academics. Since (unlike in traditional scholarship–say the study of pre-Islamic poetry in the Arabian peninsula) having written good papers is less important to GOOG than the ability to commit clear code, and since their screening process is itself derived from academic bullshit (whiteboard exams asking about algorithms & data structures), they create demand for bullshitters with certain characteristics– a pipeline of demand for big-data machine-learning bullshit. See Laszlo Bock in the NYT or Steve Yegge on his blog almost a decade ago. These do not work — but the questions are already written, and everyone else is doing it. That's my personal theory.

Further questions:

- how does university-professor recruitment like the above differ from other examples of industrial research over time — Xerox PARC, Bell Labs, JPL, Salomon, etc?

- when will this all end?

- what will the HBR write about next?

- what "actually" teaches people to think critically?

- are science and mathematics classes antithetical to critical thinking? (I notice mathematicians are incredibly bad at critical thinking. Pascal made a related comment in his 1664 memoirs. Further comments could link Soviet scientism to poor critical thinking skills.)

- why did "big data" and "algorithms" (literally translated: ways of doing things) catch on among all the other kinds of bullshit in this particular zeitgeist? (My answer: we live in a scientistic age where money and technology have replaced religion's former role, eg in providing moral guidance– see for example the prelude to "The Right to be Lazy" )

Pepper White's interesting book Learning to Think at MIT recounts that interaction with industry is what "makes" M.I.T. a success. Though engineering companies may be bullshitting themselves in this requirement, asking for a higher degree before an experienced engineer can move up in the company drives experienced engineers into M.I.T. as they try to skill up, brand themselves, and raise their salary.

The contact between professors and managers-to-be is what brings real-world knowledge into M.I.T., as well as research money/contracts to do real inventive work for major engineering firms, when the master students do get those management roles.

So, another question:

- why is the interaction between "silicon valley" (broadly construed) and academia leading to "productive" interchange? This book –out of a university press, P.U.P.– mentions a 1982 essay and I'm sure, besides Orwell's famous essay on political bullshit, we can find innumerable screeds against lying and bombast going back as far as we would like.

A salient feature of the analysis of bullshit, to me, is that Universities do put out some of the most informed, solid, truthful, and well-researched books, as well as the bullshit everyone is surely familiar with, be it from finance academics, machine-learning academics, or cultural theorists.

It would be too easy and quite wrong to say that business professors, ML professors, economists, or cultural-theory professors are full of shit–even though we can observe credible causal mechanisms and a wealth of examples of people holding such posts, who are routinely full of shit. The answer has to be more complex.

Orson Terrill writes: 

Isomorphisms asked what teaches critical thinking skills. Rigorous symbolic logic with all the proofs, like that seen in any solid discrete mathematics course, and a hard study into philosophical logic, which has much of the same content, but without flowing into set theory, and instead has a hard look at fallacious thinking via the many fallacies.

Isomorphisms replies:

A distinction between "broad logic" and proofs is an important one. Most mathematicians lack common sense (logicians and PLT theorists even more so), and most forms of argument have not been formalised within logic. For example centuries passed between St Anselm's ontological argument and Gödel's formalisation of it within modal logic. Lawvere, an avowed communist, attempted to formalise Hegel over a century after Hegel's death. If you want to argue that proofs make one wise, you'll have to contend
with the inventor of category-theory's communistic views.

The people who impress me with their critical thinking ability often display a study of history (not necessarily a college major or war buff). Some professional anthropologists–Lumonier, Malinowski, Chagnon–have impressed me with their critical thinking. (Although of course academics of any stripe often get sucked into their irrelevant peer-only backwaters. Anthropologist backwaters just happen to be leftward of business professors'.)

Method acting, like art crit, subjects the student to painful critiques of their performance. To advance a thesis, it seems like hitting people where they care–their religion (philosophy class), their pocketbook (trading), their creativity (art & acting)–making them see they were wrong where it really mattered–may be part of the key to improving students' critical thinking. The things that students don't care about–their required essay about some boring book, a proof of a fact/about an object they never inquired about–don't seem to have any impact on the core person.


Comments

Name

Email

Website

Speak your mind

2 Comments so far

  1. Andre on January 25, 2017 7:45 pm

    I think live in the midwest, then live in new york city. Then you have the bullshit/ reality spread covered.

  2. Big Al on January 27, 2017 6:18 pm

    I liked Peter Thiel’s observation that the use of the word ’science’ should alert your antennae to BS. Social science, data science etc. use the word to lend credibility in way that Physics, Chemistry etc. simply don’t have to (because we know they are right).

Archives

Resources & Links

Search