In which I wake up in a Mexican hotel

Just  spent the morning in a long rambling talk with two physicists whom I had not seen here in twenty years, namely Martin Land and Tepper Gill.  It was marvelous beyond words to see them again.

Both them remembered me from the conference in Houston in 1998. Tepper said I looked younger than then, even though I'd gone grey. I told him it must have been graduate school.

Martin and I spent about two hours talking over breakfast in the hotel before Tepper showed up. The waiter must have refilled our coffee cups half a dozen times, I barely noticed.

As is usual for physicists meeting over breakfast, we talked about everything except physics.

For example, we spent half an hour discussing our mutual interest in the Russian Formalists, as well as the linguistic theories of Roman Jakobson. Martin mentioned that his son had read Jakobson and believed he had one of the forms of aphasia that Jakobson described.

"I just read that essay!" I said.

My Elective Schedule, Humanities and All

Although my full-time preoccupation lately is barnstorming through an M.I.T.-centric computer science curriculum, I'm also following other courses, which I call "my humanities electives," although they happen to include a biology and a psychology course too.
They are:

• Yale LIT 300, Introduction to the Theory of Literature (Fry, Spring 2009)
• Yale HIST 276, France Since 1871 (Merriman, Fall 2007)
• Stanford BIO 150 Human Behavioral Biology (Sapolsky, Spring 2016)
• University of Toronto PSY434, Maps of Meaning (Peterson, Spring 2016
• Université Paris-Sorbonne, Linguistique générale (Monneret, Spring 2016)

Except for the Sorbonne linguistics course, which I picked up only recently, "with the approval of my advisor," as I joked to Red, all of these courses date from before our recent trip to France, and thus from before my rigorous comp. science "virtual degree program." 

The main difference between these and my computer science courses is that for these, I don't follow any rigid daily schedule, but instead listen and take notes at an irregular pace that suits my schedule, often in leisure. Sometimes weeks go by between lectures for any given one, often because I am doing outside reading. Nevertheless I'm committed to finishing all of them eventually.

In the case of Robert Sapolsky's excellent Human Behavioral Biology course, for example, he is such a fast speaker, and his lectures are so crammed with mind-blowing information, that I typically nibble at them, taking in only ten minutes at a time on any given session, and often looking up many references to learn details of things he talks about. 

I've already mentioned Paul Fry's Theory of Literature course, which has changed my life, and for which I am doing a huge amount of outside reading. 

The other Yale course, taught by John Merriman is the one that kicked off this whole revolution in my life of following online college courses, as much like a student as possible. That being said, I'm only a third of the way through the course, as I've suspended watching new lectures until I finish reading Germinal by Emile Zola, since it's one of the main reading assignments for the course.

From that course, I quickly saw that Yale had tremendous online offerings that included the lecture notes. They are probably second only to M.I.T. in terms of having their act together this way (and both are far ahead of Harvard, for example).

Maps of Meaning is somewhat of an outlier in the list. Jordan Peterson has been so much in the news lately, often in controversial ways, given the ideology of the media class. Just mentioning his name probably makes some folks fly into apoplexy. I had already seen many of his shorter videos and lectures, and I wanted to "sit it on" one of his real classes. 

I have been well rewarded for this, and have enjoyed the lectures very much (I am almost done with the course). They are fascinating mix of psychology, neurology, evolutionary theory, religion, mythology, history, philosophy and literature. He will go from talking about brain functioning to Jung, Nietzsche, the Soviet gulag, the Egyptian creation story, all over the course of twenty minutes. Yet he cycles back around to the same subjects in a way that makes it coherent over time.

The difference between his course and the others, however, is that Peterson's lectures are home-brewed*, and apparently not at all assisted by anyone official on the university level (if you knew Peterson's troubles with the U of T, you might not be surprised at this). He sets up a camera and records himself, then uploads the video. This means, at least in the 2016 course, we rarely see what he is talking about on the projector screen. 

Also there are parts missing from some of his lectures. That brings to mind the funny fact of how I used to not care so much about missing class when I was younger. Now it's a huge deal to miss out, and I feel massively deprived in such a case. 

Last week I found out that the last fifteen minutes of Lecture 11 of my M.I.T. mathematics course was chopped off. Van Dijk just left off in the middle of a proof on partially ordered sets (something that comes up in computer modeling of task scheduling).  Several other folks had already noticed this, and the M.I.T. Youtube account folks apologized for this in the comments to the video. 

Fortunately I was able to go the typeset course lecture notes on the M.I.T. website and recreate the note-taking experience, pretending that I was narrating the material in Dr. Van Dijk's patient patter until I had everything down as I needed.

*Edit: I just realized that, of course, Monneret's linguistics course is probably home-brewed tube. One only gets the audio, which is ok because he speaks clearly enough for me to understand most of the time, and he has a separate site where one can see all the slides he is using. In a way, that's better than regular video. 

Ironically (since it's a linguistics course) one of the funniest things about the lectures is the automatic captioning of his lectures into French, which at times butchers the French into nonsense. I'm not surprised at this. Not only is French arguably a harder language than English to transcribe using Speech-to-Text (because of having fewer phonemes at play in its core vocabulary), but also because the A.I. at work  here has had, I would suppose, much fewer "good" examples to work with, and is thus less accurate. That's how A.I. works. It learns by being taught.

Virtual M.I.T.

Today was the last day of the regular weekday M.I.T./Stanford class schedule I've following. For the last month, Monday through Friday I've been pounding away at this:

both:

• MIT 6.00 Introduction to Computer Science and Programming (Grimson and Guttag, originally taught Fall 2008)
• MIT 6.002 Circuits and Electronics (Agarwal, Spring 2007)

and on alternating days, either of these:

• MIT 6.042J Mathematics for Computer Science (Leighton and Van Dijk, Fall 2012)
• Stanford CS107 Programming Paradigms (Cain, 2012)

By "pounding away," I mean watching each lecture on Youtube, and taking detailed notes. 

It's the last day of this particular schedule, because I finished the 6.00 course (24 lectures in all). 

I've still got 4 lectures to go on Circuit, and I'm only half way through the Math course, and one third of the way through the Stanford course. I'll continue with those up after a week's break.

I'm going through a couple other M.I.T. comp sci courses besides these, but I've put the others on hold until I finish these.

The 6.00 course was fantastic. Absolutely worth it. Although it was introductory, I still learned quite a bit, and it was great to formalize the knowledge I already had. Professor Guttag wrapped it up today with a nice review and summary that featured a description of his own research, as an example of applications of computer science. 

Among other things, his research group was (as of 2008) using applied machine learning to build seizure warning devices for epileptics that recognize variations in the brain patterns of individuals (currently a difficult problem), and similar A.I.-based implanted defibrillators for people with cardiac disease. Fascinating!

His teaching partner, Eric Grimson, is one of the big researchers using A.I. to develop advanced medical imaging.

All of the professors I have for these courses (like Paul Fry for my Yale lit course) are very inspiring. Tom Leighton, for example, founded Akamai, a company that runs the servers that power a good chunk of the Internet, including (at least at one time) Facebook.  He was able to develop mathematical algorithms at M.I.T. that allowed for efficient distribution of load balancing of a gazillion web requests, and leveraged this into a fortune.

Likewise Anant Agarwal turns out to have founded the online learning platform edX. I had no idea until about week ago that his old Circuits course that I've been following has evolved into  one of the flagship courses on edX. He's a huge reason why M.I.T. is the head-and-shoulders leader of online web-based learning on the university level. Just one more amazing thing that has come out of that place.

The Transitioning of Shadows Before Dawn

Rising before dawn, one sees the bright moonlight differently then in the evening after sunset, which is how I have seen it most of my life.

When one expects deep darkness at an hour, it is startling, even shocking, to wake up to the delineated contrast of the milky glow on the patio pavement alternating with the deep shadows cast by the side of the house..

In modern times, with our typical hours, we tend to experience the emergence of the full moon from its playful debut on the horizon in gradual increments of increasing radiance towards its solo performance is blazing whiteness against the night sky..

To view the reverse outside my window as the dawn approaches---the powerful orb on dark blue sky losing its shadow-casting power while still so bright---is a curious phenomenon that captures my attention, and distracts me from an attempt to read an essay on French poetry before breakfast.