Sunday, 10 January 2021

Perhaps a Different Perspective on Online Teaching

In March last year, the pandemic dramatically struck close to home when universities across Canada (and probably in many places around the world) abruptly moved from delivering lectures in person to online. The suddenness with which it happened meant there was scant preparation, with huge impact on professors/instructors and students. In general, the move went well -- a testament to everyone working together to make the best of an unfortunate situation.

Come September, little had changed and many universities, including the University of Alberta, conducted most classes online. This continues for the term that is just starting.

I taught a course on Operating Systems twice last year: January-April (Winter 2020) and September-December (Fall 2020). My Winter 2020 term started off in the classroom with me using PowerPoint slides. This was supplemented with numerous interactive exercises that used the classroom whiteboard, as well as a "human animation" example. For Fall 2020, it was PowerPoint over Zoom, with no whiteboard tools. (Yes, there are whiteboard tools for online teaching, but they are a poor match for my needs.) 

Having been out of the classroom for more than a decade, in Winter 2020 I was disappointed to see that the trends emerging in the early 2000s had continued. I had seen class attendance slowly decline and students in class become distracted by their laptop and/or cell phone. The number of in-class questions kept dropping as compared to what I remember from the pre-social media days. Few students were willing to speak up in class and answer questions that I posed. In effect, the class was largely a one-way communication: I was a talking head at the front of the room.

I had more fun teaching my Fall 2000 course than I have had in many, many years (118 students).

It took a couple of weeks, but the students warmed to using chat. I ended up going more slowly through the material for each lecture because of the interesting questions being asked, often anonymously, through chat. I stayed online after class to continue a dialogue with engaged students (often over 40!) and that sometimes lasted 30 minutes. I've found that being online (and possibly being anonymous) has dramatically increased the willingness of students to interact with me. The students have made Operating Systems much more interactive than I ever imagined it would be, more engaging and educational for the students, and more fun for me. A winning situation for everyone!

The only downside is that I covered the course material slower than I anticipated (because of the in-class questions) and had to rush through the last few lectures to make sure all the material was presented.

Classes start on Monday for the Winter 2021 term. I am excited to be back in the (virtual) classroom and look forward to engaging with the students in a meaningful way to enhance their learning experience.

Details: Students could ask a question through chat by sending it to the entire class or privately to me. That meant that students afraid of asking questions publicly could still direct a question to me during the class. This was valuable because, quite frankly, these students often asked "I don't understand this. Can you re-explain it?" -- something they felt intimidated about doing in a public setting (but usually did after class when no one was around to witness their "ignorance"). I respected the student's privacy by reading the question asked out loud, but without the student's name. By verbalizing all the questions, there was no need to also publish the chat window -- the classroom video was self-contained.

Saturday, 9 January 2021

The Longest Ongoing Experiment in Computing Science History

On August 31, 1970 a historic computer science event began. It happened in a downtown New York hotel as part of an academic conference. The event spanned three days, after which the participants returned home to various places in the United States. The event was billed as the First United States Computer Chess Championship, a seemingly futuristic competition that pitted computer against computer at the human game of chess. To some, it was humorous entertainment; to others it was a glimpse into the future potential for artificial intelligence (AI) technology.

The event attracted a smattering of media attention, but was novel enough that there was a quick decision to repeat it the following year. And the next year, and the year after that, and… The experiment that began in 1970 continues to this day.

Today, we can look retrospectively at fifty years of computer chess competitions and applaud the amazing progress that has been made. This includes advances in algorithms, hardware design, and software tools have allowed the AI community to make astonishing leaps forward, for chess in particular and games/puzzles (one-player games) in general. The work in using games for AI research had numerous benefits: ideas that propagated to non-game applications (such as transposition tables, iterative deepening, and Monte Carlo tree search); use as a pioneering experimental test-bed (e.g., for reinforcement learning, planning, and optimization); and as public demonstrations of the potential for superhuman performance (e.g., checkers, chess, go, and poker).

I was pleased to organize an effort to publish a special issue of the International Computer Chess Association Journal devoted to the first 50 years of computer chess tournaments, beginning with the first one in 1970 and continuing to the present day. Most of the contents are historical papers that help document the early days of computer chess. This is an important initiative to be doing now while many of the people who created the history are able to give their first-hand perspectives.

The papers in this issue are as follows:
  • I write about the 1st United States Computer Chess Championship, held in New York in 1970. Enjoy the story of this historic event.
  • David Slate co-authored Chess 3.0 (with Larry Atkin), the winner of the 1970 event. The Chess x.y series of programs dominated computer chess in the 1970s, including winning the 1977 World Championship. He tells us about the 1976 Paul Masson human chess tournament and the challenges he faced being the sole computer entry.
  • Tony Marsland relates stories and events in his long computer chess career, including the Marsland CP’s participation in the 1970 event (his program was later called Wita and then Awit).
  • Monty Newborn proposed the 1970 event to the Association for Computing Machinery (ACM), organized the 1970 event, and was a frequent participant in the 1970s with Ostrich. His article analyzes the game results between the top programs over the past 50 years.
  • David Levy was guest commentator and a participant at many of the ACM computer chess tournaments. He tells us about the important role that Benjamin Mittman played in computer chess and the International Computer Chess Association.
  • Linda Scherzer reminisces about the chess machine Bebe, three-time runner-up in the World Computer Chess Championships (1983, 1986, 1989).
  • Robert Hyatt talks about the evolution of Blitz to Cray Blitz (adding a supercomputer) to Crafty (open source). Cray Blitz won the 1983 and 1986 World Computer Chess Championships.
  • Ulf Lorenz discusses the history of Hydra, a specially built chess machine that dominated computer chess in the early 2000s.
  • Ingo Althöfer provides the little known history of computer chess in East Germany – before the wall came down in 1989.
Members of the International Computer Games Association (ICGA), formerly the International Computer Chess Association (ICCA), should be proud of the role that this organization played in making artificial intelligence and computer science history. Our contribution was through organizing annual computer chess tournaments, computer Olympiads (for a wide variety of games), research conferences, and the flagshipInternational Computer Games Association Journal.

Wednesday, 11 November 2020

Computer Olympiad 2020

In August 2019, I was proud to be appointed the President of the International Computer Games Association (ICGA). The ICGA started out in the late 1970s as the International Computer Chess Association (ICCA) to promote the development of AI in game-playing programs. The organization changed its name roughly 20 years ago to reflect the AI community's interest in a wide spectrum of games.

The ICCA/ICGA has three events to promote AI in games. Most are held annually.
1) The World Computer Chess Championship. First held in 1974, these events bring the strongest chess programs together to compete for the ultimate title. I co-organized the 1989 event in Edmonton -- Deep Thought (the predecessor of Deep Blue) was the winner.

2) The Computer Olympiad. Started in 1989, this event is the "Olympics" of computer games. In a given year there will be 10-20 tournaments in a wide variety of games. New games are continually added to the list, and some games drop off (once they are solved).

3) In alternating years, the Computers and Games and Advances in Computer Games conferences are held. The ICGA has been holding conferences/workshops for over 30 years.

In addition, the ICGA produces the quarterly ICGA Journal, containing refereed articles as well as tournament reports and news stories relating to games (over 35 years and counting).

For me, personally, the ICCA/ICGA has been tremendous for my career. Some of my earliest academic papers appeared in the ICCA Journal. Starting in 1980, I began competing in computer chess championships, including tying for first place in the 1986 World Championship. The checkers program Chinook's first tournament was in the 1989 Computer Olympiad (1st place and the gold medal). The success in this event was pivotal to the future of this important project.

Along the way, I have met many talented people in the computer games and AI world through my association with the ICCA and ICGA. Some of my fondest academic memories occurred at ICCA/ICGA events.

====

The pandemic disrupted this year's ICGA events. Our conference and the World Computer Chess Championship were cancelled. However, the Computer Olympiad is a go! November 23 is the deadline for registration. Here's your chance to compete for a gold medal. Check here for more information. Good luck!

Friday, 11 September 2020

Market of Monsters -- Complete!

A proud father says congratulations to Rebecca Schaeffer! The final book in her Market of Monsters trilogy is now available. She started working on the project in the summer of 2015, landed a publishing deal in late 2016, and the first book (Not Even Bones) appeared in September 2018. Now this highly-rated young-adult horror series is complete (no spoilers). Not Even Bones will hook you with its originality and cleverness, and force you to compulsively devour the book because of the real/implied horror. If you liked the television series Dexter, you will love the Market of Monsters trilogy.

Book 1 (September 2018)

Book 2 (September 2019)

Book 3 (September 2020)

Tuesday, 17 September 2019

World Checkers Champion -- 25 years later

August 18, 2019, was the 25th anniversary of a personal and professional milestone. On that day in 1994, the University of Alberta program Chinook won the World Man-Machine Checkers Championship. Chinook was the first program in any game to win a human world championship, a feat recognized by the Guinness Book of World Records.
Chess grandmaster Raymond Keene (left) speaking at the opening ceremony of the World Man-Machine Checkers Championship, with Jonathan Schaeffer (representing Chinook) and Marion Tinsley (World Champion)
The victory was bittersweet. After Chinook narrowly losing the Championship in 1992 to Marion Tinsley, I spent most of the next two years preparing for the rematch. Then, in Boston, after 6 games of a 20-game match, all draws, Tinsley uttered the words that haunt me to this day:

“I resign the match and the title to Chinook.”

There is no need to go into the details of what happened next and why; they are recounted in my book One Jump Ahead (Springer-Verlag, 1997 and 2009). Suffice to say that Tinsley felt ill, and it was unthinkable to continue playing against an opponent who could not perform at their highest level. Chinook was awarded the match on forfeit and subsequently defended the title twice. Along the way there was much animosity expressed by a few people external to the match, taking a bad situation and turning it into an awful mess. It didn't end there -- less than a year later, Tinsley passed away from the disease that manifested itself at the 1994 match.
Green Lawn Cemetery, Columbus, Ohio -- note the position on the checkerboard, taken from a Chinook-Tinsley game
In 1997 Chinook was retiredhaving vanquished all comers.

Rob Lake, a member of the Chinook team, reminded me of this milestone a month ago. So, why has it taken me so long for me to post something? I am immensely grateful to have been part of an amazing and historic adventure, and I treasure my memories of the amazing Marion Tinsley. However, the events of 1994 were painful. It took an enormous effort in 1996/1997 for me to put the 1994 story into words for One Jump AheadTime has done little to diminish the discomfort. I find it more comfortable to contemplate the future rather than dwell on the past.