On Friday and Saturday I had the exceptional opportunity to attend the ACM Turing Centenary Celebration. I’ll start right off by saying it was an exciting and deeply intimidating event to attend. The ACM (Association for Computing Machinery) is actually quite an academic-focused organization in general, and while I read their Communications of the ACM magazine, I find myself struggling through some of the articles every month. At the conference itself, not only was I surrounded by 32 Turing award laureates (the best computer minds in the world) it seemed like a vast majority of the attendees were from colleges and universities as professors, researchers or students (both undergrad and graduate level). And me? I never went to college and I work in the industry doing systems administration. I quickly felt like a mere technician in this crowd. That said, the reasons for it being an exciting event were very similar, I was in the same room as Vint Cerf, Donald Knuth and Ken Thompson! My computer heroes!
The event schedule, full program and videos are all available if you’re interested in more comprehensive covering of the sessions, as I’ll just cover some of the highlights for me from the 6 pages of notes I took while I was there.
I was quite fortunate that the venue was not only in San Francisco, but it was a mere 2 blocks from where I live so I headed out around 8:30AM to find a seat for the 9AM start of the event. Opening remarks were made by John White and Vint Cerf.
Vint Cerf giving opening remarks
The first panel was titled Turing the Man. The panel had people who knew Turing or were intimately familiar with him and his work on some kind of personal level and the charming stories they told really set a good tone kicking off the conference. Next up was the Human and Machine Intelligence panel where they discussed some of the history of Artificial Intelligence, starting from Turing’s work on the subject and leading into some of the latest developments with Watson and Siri. They went through the basic definition according to Turing and his “imitation game” (or as it’s known now, Turing Test), which focuses on behavior of intelligence, rather than any kind of philosophical or religious definition of intelligence, and which builds the foundation of AI work today. They reviewed some of the key tenants of AI, covering assumptions about the world that AI scientists have to consider and the technical side of how to implement things like storage and search capabilities required for AI today and some of the challenges they continue to encounter. There was also talk about how AI can be used in searches to intelligently narrow down sections of material to do more precise searches on (an example was medical records). AI is one of those topics that make me wish I’d taken an academic route in life, the subject is infinitely interesting to me and I really enjoyed the positiveness and excitement of this panel.
AI Panel
Before lunch Butler Lampson gave a talk on What Computers Do: Model, Connect, and Engage where he covered key portions of what computers have done since their inception and what we’re looking forward to in the future when it comes to engagement in the physical world. Already we’re seeing some very interesting advances from the Microsoft Kinect technology which has already been applied far beyond the gaming use it was built for, the advances being made in self-driving cars and work being done with augmented reality. He also reviewed some of the challenges with all this new technology, including fault tolerance, agility, scalability and dependability.
Butler Lampson on What Computers Do
Since I didn’t know anyone at the conference lunch was a bit awkward at first, but I ended up talking to a woman from IEEE who, upon learning of my background and current work, was interested in my perspective on the practical usefulness of some of the articles in one of their publications. We exchanged contact information so hopefully I’ll be doing that review pretty soon.
The first panel after lunch was Systems Architecture, Design, Engineering, and Verification – The Practice in Research and Research in Practice where several Turing award winners discussed the work they had done that not only had significant impact on research, but were directly applicable in practice. The talks covered some history, their feelings on model checking and formal verification of industry software, some infamous software bugs and what led to them and more general rigorous software design. There was also discussion on what steps need to be taken to create a more stable software future, including increased collaboration between people who are not only able to see the practical concerns of security and proper testing methods, but also who can implement them well. Part of the panel also dipped into talking about where we see computing going in the future, including a speculation about terminal-based computing (no more desktop, just a screen that hooks into a network) and related concerns about data security and privacy in a network “cloud” based world.
The next talk was given by Alan Kay on Extracting Energy from the Turing Tarpit. The Turing tarpit is a phrase defined by Alan Perlis statement “Beware of the Turing tar-pit in which everything is possible but nothing of interest is easy.” Kay presented a different way of approaching this which turned the problem into more of an opportunity. But I think what fascinated me the most about this talk was learning about Sketchpad, which I’m somewhat embarrassed to admit I had never heard about, but was a revolutionary program for 1963 and helped lay the groundwork for not only CAD software of the future, but graphical interfaces in general, very cool. Following a break it was on the another panel, The Turing Computational Model and How it Shaped Computer Science. This is where the day started to get hard for me, my knowledge of the Turing Computational Model was not sufficient to do more than tangentially follow along with this talk, I’ll be doing some reading and hope to re-watch it when I’m better prepared. As someone whose math education ended with Functions, Statistics and Trigonometry (the class before Precalculus in my high school), the last two talks of the day were beyond me, the Lambda Calculus Then and Now talk being completely out of my grasp and The Computable Reals and Why The Are Still Important being one I only really managed to follow portions of. Phew!
The next day started bright and early at 8:30AM with a Computer Architecture panel. The panelists discussed some of their own work, including Ivan Sutherland’s work and predictions related to clockless computers (a Scientific American article “Computers without Clocks” gives a nice, human-readable version of his work). Fred Brooks went on to discuss how Turing’s hardware proposal of the Automatic Computing Engine weren’t actually all that influential compared to his peers due several factors, including coming a bit too late, not being collaborative enough in the work, having some peculiarities and, most telling, that he didn’t realize how important ease of programming would be for other people. The panel wrapped up with some discussion on how important a core of Math and Physics are in computer science, and a lament that there are many programmers who don’t have a fundamental understanding of how computers work, which impacts their ability to write efficient code. The next panel was Programming Languages – Past Achievements and Future Challenges. The panel began with a retrospective look at early work in cryptography and led into a history of how programming languages and compilers quickly became a part of Computer Science. There was a fair amount of discussion about how complicated the really good programming languages are and how this has led to easier, less powerful, languages being used it production and consensus from the panel that this was not a great direction to be going in. Barbara Liskov mentioned that she hoped a future language could provide the ease of writing and the power and important programming concepts required for really effective programs. Simplicity certainly was an underlying current in this panel, and there was some talk of domain-specific languages having a future because they offer the ability to reduce complexity.
Programming Panel
Then it was on to a panel on The Algorithmic Universe. One of the first observations from this panel was from Don Knuth who discussed the percentage of people who were “algorithmic thinkers” and how he believes this is increasing due to the needs of our modern world, and later they agreed that this type of thinking could be taught. Much of the panel focused on how algorithms are being used by natural sciences and how we need to make sure we don’t have a narrow view of what “computation” is when doing work beyond the Computer Science realm. The last panel of the conference was one of the most entertaining, Information, Data, Security in a Networked Future. They covered brokenness of user passwords for security, desire for universal identification systems and the failures of SSL (we all want PKI to work, but this is a major example of a PKI that has a lot of failures). They also covered a common failure with secure systems going back at least as far as cracking Enigma codes in WWII: overestimating the user, underestimating the computing power of an attacker. The talk then went to some of the politics that have been getting into the computing realm, and the panelists suggesting that more effort be put into groups who can educate the governments and officials who make decisions, or having computer scientists become some of these decision makers. There was also talk of social networks, concerns about popular ones currently being controlled by companies who have little to no assurances of data retention or portability and the consequences of their use not only by the people actively using them and their immediate social groups, but for anyone within camera range in public today.
The conference officially adjourned at this point, but many (myself included) stuck around for a screening of the short version of Codebreaker. As a documentary geek, I thoroughly enjoyed it, even if it was quite sad. I look forward to seeing the full version when it’s released on DVD next year.
I’ve uploaded more photos I took during the conference here: http://www.flickr.com/photos/pleia2/sets/72157630140487140/
I’m very glad I was able to attend, it was quite an amazing opportunity and I learned a lot.