Artificial Intelligence and Machine Learning BLOG@CACM

How WWII Was Won, and Why CS Students Feel Unappreciated

The Communications Web site,, features more than a dozen bloggers in the BLOG@CACM community. In each issue of Communications, we'll publish selected posts or excerpts.

Follow us on Twitter at

John Arquilla considers how code-breaking helped end a war, while Jeremy Roschelle ponders the use of music in data science education.
  1. John Arquilla: Hacking the Axis
  2. Jeremy Roschelle: Learning Computational Thinking to Dominate the Music Industry
  3. Authors
BLOG@CACM logo May 7, 2020

Observations of the 75th anniversary of the end of World War II in Europe (May 8, 1945) included remembrances of such searing events as the struggle on Omaha Beach on D-Day, the Battle of the Bulge, and at least some recognition of the enormous contribution made by the Russian people to the defeat of Fascism. Yet in all this, I suspect the role of the first "high-performance computing" capabilities of the Allies—known as Ultra in Britain, Magic in the U.S.—will receive too little attention.

The truth of the matter is that the ability to hack into Axis communications made possible many Allied successes in the field, at sea, and in the air.

Alan Turing and other "boffins" at Britain's Bletchley Park facility built the machine—a much-improved version of a prototype developed by the Poles in the interwar period—that had sufficient computing power to break the German Enigma encoding system developed by Arthur Scherbius. The Enigma machine was a typewriter-like device with three rotors, each with an alphabet of its own, so each keystroke could create 17,576 possible meanings (26 x 26 x 26). When a fourth rotor was added, the possibilities rose to 456,976 per keystroke.

The Germans had faith in their system, but Turing & Co. met and mastered this challenge. The timely information they decrypted had profound effects at many critical moments. When Erwin Rommel and his Afrika Korps made their final lunge toward the Nile, Ultra intercepts kept the British informed of his exact plan of attack—for which they prepared well, then repulsed. In the Battle of the Atlantic, Ultra hacks not only allowed for the rerouting of convoys away from these predators, but also enabled subhunters to turn up and attack U-boats and their supply ships at even the most remote ocean locations.

Much Ultra-hacked information was shared with the Russians, too—to some extent under cover of a "legend" that the secret material was being provided by a British-run human spy ring. This proved crucial in many Eastern Front actions, but most notably in the massive tank battle at Kursk in July 1943, which truly broke the back of Hitler's panzers. At this point, the Germans became convinced some traitor was leaking their most highly classified information to the Allies, but they never lost faith in Enigma.

Nor did the Japanese ever give up on their Imperial Codes, the Magic hacking of which led to the ambush of Admiral Yamamoto's massive forces at Midway, and greatly informed the American sub-marine campaign against Japanese shipping. U.S. Navy "pigboats" sank over 80% of Japan's merchant ships, and about one-third of the Imperial Navy's warships, almost always guided by Magic hacks. Indeed, the level of detail was so great that, in all the vast Pacific, an American submarine commander often had such exact information that he knew enemy ships' names, cargoes, even what the noon position of the ship would be on its course the following day!

Truly, the impact of this first "information war" was profound. Had the Axis powers been less complacent about the robustness of their codes, the outcomes of critical battles and campaigns could well have gone in their favor, rather than against them. The lesson for today from this very cautionary tale is that the cybersecurity of armed forces is absolutely crucial to their physical security, and to their prospects for victory.

So, on this 75th anniversary of a war best known and remembered for its range of startling new weapons and the sheer grit of its soldiery in battle, let us take just a moment to recognize the pioneering high-performance computing capacity of the Allies contributed most significantly to the final margin of victory.

Back to Top

Jeremy Roschelle: Learning Computational Thinking to Dominate the Music Industry April 22, 2020

One of my early experiences in computing involved using the Music Logo programming language to "program" something that sounded like Beethoven's Fifth Symphony. Working with MIT professor Jeanne Bamberger (, I used music as a context for programming and for inquiry into music. This process reshaped my views of what computing could be. I came to see computer science as providing a frame of analysis that could reveal the internal structure and patterns with-in a wide variety of human experiences, not just those that intrinsically involve computers. This experience led to my career as a learning scientist.

More recently, I have had the opportunity to use music as a context to get middle school students in New York City excited about data science. I served in a consulting role for a multi-institutional team that developed Beats Empire, a game in which a student manages an artist's rise to fame and fortune. To help their artists, the students must demonstrate what they are learning about using data to analyze music industry trends. The game both engages students and can give teachers a sense of what students know and can do. It is available for free at

Here I will share thoughts about music as a context for learning about computing.

  1. Music as an authentic, accessible context

Students have complex experiences with music. They don't just listen to music; they talk about how artists use social media. They discuss streaming services and how they recommend music. They think about themes in song titles and lyrics. They think about who listens to music where and on what devices. This rich encounter of music on computing platforms can set the stage for a learning opportunity where students go behind the scenes to see how computing influences our experience of music.

In interviewing music industry experts, we found experts could easily and cogently explain to students how data is being used to shape artist's music and careers—and why data science in the music industry can be a great career for women and people of color. For example, see the interview "A Visit to Chartmetric" (, where I visited a company that specializes in creating analytics dashboards for artists and their managers. Chartmetric was willing to explain what they do and why they love their jobs to middle school students.

  1. Students' experience of a drive for data

As a learning researcher, I have been involved in many projects that try to involve students with realistic data. Unfortunately, as educators we often come up with "authentic" contexts that aren't really something students ordinarily would do. In math, I know I have created a "manage a soccer team" unit where students looked at data about how fast team members can run a dash.

One thing I have learned is that in a game context, one can simulate a role where data collection is not "assigned" to students, but where they start from a purpose they care about: helping an artist grow their career. In Beats Empire, students sometimes make spontaneous decisions for their artists. For example, they can recommend a mood or theme for an artist's song based on their intuition. But they can also collect data in the game, for example on trends in moods and themes. They look into what is popular in particular neighborhoods via an in-game map. The game is set up so paying attention to data and trends can dramatically increase the success of the player's artist. This creates a relationship to data that is much more like the real world; data is not as a context for a specific math concept or science principle, but rather as a tool for getting better at what you care about—in this case, music. This can be exciting to students.

  1. Students' opportunity to iterate with data

It's also very common in a math or science class to cycle through using data only once. In a science lab, you collect the data, analyze it, report it, and you are done. In this context, the coherence between the processes of collecting, storing, analyzing, and interpreting is often only in the eyes of the curriculum designer or teacher, who make sure the phases of the cycle fit together. But students need to learn this, too.

It is important for students to see how the separate processes of collecting, storing, analyzing, and interpreting data constrain each other. If you want to analyze a trend of a particular kind, it's important to collect and store the data in an appropriate way. A gaming context can create a situation where iteratively looping through processes happens quickly and is essential to the game play. Students can start to connect their choice of analysis types to actions they can take in the game. Likewise, they could decide how to collect data based on the questions they want to answer. There are important lessons to be learned in how to iteratively refine the relationship between data and an overall purpose or initiative.

Students are always learning. Meet them where they are.

Overall, the field of the Learning Sciences recognizes learning is not a special type of experience that only happens in designated settings. People are always learning. Too often, we begin by thinking about what students are not learning and then we try to create an artificial experience so they will learn. But it's also possible to take a context in which many students are enjoying learning about every day, like music, sports, food, or fashion, and think about how to deepen their learning about that experience. One path is by layering computing-rich experiences into the contexts students are already motivated to learn about. Games can provide a bridge between what students like to learn and enhanced opportunities that will lead towards a career in computing.

This material is based on work supported by the National Science Foundation under Grants No. 1742011 and 1741956. Opinions, findings, conclusions, or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the National Science Foundation.

Join the Discussion (0)

Become a Member or Sign In to Post a Comment

The Latest from CACM

Shape the Future of Computing

ACM encourages its members to take a direct hand in shaping the future of the association. There are more ways than ever to get involved.

Get Involved

Communications of the ACM (CACM) is now a fully Open Access publication.

By opening CACM to the world, we hope to increase engagement among the broader computer science community and encourage non-members to discover the rich resources ACM has to offer.

Learn More