So You Want to Learn About Quantum Computing?

“Students will learn by inhabiting an alternate history where Alan Turing and Richard Feynman meet during World War II and must invent quantum computers to defeat Nazi Germany. As a final project, they will get to program a D-Wave One machine and interpret its results.”

If you are based in Seattle then you want to keep an eye out for when Paul Pham next teaches the Quantum Computing for Beginners course that follows the exciting narrative outlined above.

For everybody else, there is EdX‘s CS191x Quantum Mechanics and Quantum Computation course.  I very much hope this course will a be a regular offering.  Although it lacks the unique dramatic arche of P.Pham’s story line this course is nevertheless thoroughly enjoyable.

When I signed up for this course, I didn’t know what to expect.  Mostly, I decided to check it out because I was curious to see how the subject would be taught, and because I wanted to experience how well a web-based platform could support academic teaching.

This course fell during an extremely busy time, not only because of a large professional work load, but also because the deteriorating health of my father required me to fly twice from Toronto to my parents in Germany.  Despite this, the time required for this course proved to be entirely manageable.  If you have an advanced degree in math, physics or engineering, and want to learn about Quantum Computing, you shouldn’t shy away from taking this course as long as you have an hour to spare each week.  It helps that you can accelerate the video lectures to 1 1/2 normal speed (although this made Prof. Umesh Vazirani sound a bit like he inhaled helium).

Prof. Vazirani is a very competent presenter, and you can tell that a lot of thought went into how to approach the subject, i.e. how to ease into the strangeness of Quantum Mechanics for those who are new to it. I was suspicious of the claim made at the outset, that the required mathematics would be introduced and developed as needed during the course, but it seems to me that this was executed quite well. (Having been already familiar with the required math, I don’t really know if it’ll work for somebody completely new to it, but it seems to me that indeed the only pre-requisite required was a familiarity with linear algebra).

It is interesting to see discussions posted by individuals who took the course and were apparently subjected to QM for the first time.  One such thread started this way:

“I got 100. It was really a fun. Did I understand anything? I would say I understood nothing.”

To me this illuminates the fact that you simply cannot avoid the discussion of the interpretation of quantum mechanics.  Obviously this subject is still very contentious, and Prof. Vazirani touched on it when discussing the Bell inequalities in a very concise and easy to understand manner.  Yet, I think judging from the confusion of these ‘straight A’ students there needs to be more of it.  It is not enough to assert that Einstein probably would have reconsidered his stance if he knew about these results.  Yes, he would have given up on a conventional local hidden variable approach, but I am quite certain his preference would have then shifted to finding a topological non-local field theory.

Of course, there is only so much that can be covered given the course’s duration. Other aspects there were missing: Quantum Error Correction, Topological and Adiabatic Quantum Computing and especially Quantum Annealing.  The latter was probably the most glaring omission, since this is the only technology in this space that is already commercially available.

Generally, I found that everything that was covered, was covered very well.  For instance, if you ever wondered how exactly Grover’s and Shor’s algorithms work, you will have learned this after taking the course. I especially found the homework assignments wonderful brain teasers that helped me take my mind off of more worrisome issues at hand.  I think I will miss them. They were comprised of well thought out exercises, and as with any science course, it is really the exercises that help you understand and learn the material.

On the other hand, the assignments and exams also highlighted the strengths and weaknesses of the technology underlying the courseware.  Generally, entering formulas worked fine, but sometimes the solver was acting up and it wasn’t always entirely clear why (i.e. how many digits were required when giving a numerical answer, or certain algebraically equivalent terms were not recognized properly).  While this presented the occasional obstacle, on the upside you get the immediate gratification of instance feedback and a very nice progress tracking that allows you to see exactly how you are doing. The following is a screenshot of my final tally. The final fell during a week in which I was especially hard pressed for time, and so I slacked off, just guesstimating the last couple of answers (with mixed results).  In comparison to a conventional class, knowing exactly when you have already achieved a passing score via the tracking graph makes this a risk- and stress-free strategy.

Screen Shot 2013-04-27 at 11.56.31 AMA common criticism of online learning in comparison to the established ways of doing things is the missing classroom experience and interaction with the professor and teaching staff.  To counter this, discussion boards were linked to all assignments, and discussion of the taught material was encouraged.  Unfortunately, since my time was at a premium I couldn’t participate as much as I would have liked, but I was positively surprised with how responsive the teaching assistants answered questions that were put to them (even over the weekends).

This is all the more impressive given the numbers of students that were enrolled in this course:

The geographic reach was no less impressive:

Having being sceptical going into this, I’ve since become a convert.  Just as Khan Academy is revolutionizing the K12 education, EdX and similar platforms like Cousera represent the future for academic teaching.


10 thoughts on “So You Want to Learn About Quantum Computing?

    1. Thank you, Bob.

      Unfortunately my dad is suffering from Alzheimers, so it’s a one way street.

  1. > The final fell during a week in which I was especially hard pressed for time, and so I slacked off, just guesstimating the last couple of answers (with mixed results).

    I also had a perfect mark going into the final but my “regular” school exam week coincided. So I basically did the same, made sure I answered the first few correctly, got the 80% to pass, then rushed through with some “educated” guesses to finish off the rest a few hours before deadline. I didn’t give the final the effort it deserved.

    1. Steve, I share this regret. It would have been neat and entirely doable to get a perfect score, but in the end this is about what kind of learning you derive from it, and so I tried to leave the problems that were least interesting to me. It’s nice that the individual progress report allows you to play it that way.

      For instance the question as to how Grover’s algorithm progresses if not terminated at the optimal point. Due to the unitary nature of the evolution it was clear that the system would revisit earlier states, and doing the math would have been just an add-on to the earlier assignment work. On the other hand this allowed to just fill in some results from earlier problem assignments, and to see if one took, before the numbers of tries were exhausted (i.e. I managed to guess one more right answer that way).

  2. Thanks for sharing your MOOC experience! I am a physicist, too, and a life-long moonlighting physics learner and enthusiast, but I was skeptical so far about MOOCs in physics. I was not sure if and how the ‘traditional way to learn physics’ – that is: by solving lots of problems – can be successfully transferred to an online learning setup.
    But your post really motivated to me sign up for a MOOC someday.

    1. Your comment prompted me to look what’s out there in terms of advanced physics courses, but at this point there isn’t really anything that catches my eye, although the technical plattform is certainly good enough at this point.

      I came across this little jewel: A video lecture set of Feynman talking about QED to a general audience. Graduate level lecture videos are harder to come by (although there are some). I wonder if the “market” for these more advanced courses is too small for free MOOC offers.

      1. Thanks – and what a coincidence. Yes, I know this Feynman lecture – awesome – and it was exactly David Tong’s course I had picked as my QFT primer some time ago as lecture notes are available as well on his web site.
        Although without any interaction this doesn’t count as a MOOC, correct?

        There are lots of other interesting courses on Perimeter Institute’s video server – I have spotted some on quantum computing as well. I found it just a bit difficult using that search interface to find courses (versus single lectures).

        1. It’s a small world web when looking for something that specialized. Looks like Tong’s course is the next best thing to a MOOC at this point.

          It would be great if Perimeter could team up with EdX or Cousera to build on this material.

Comments are closed.