The second person to adopt Scrum for his company was Mike Beedle in 1994. He had been following the creation of Scrum in the internet newsgroups and said, “I get it, it is all based on AI!” Mike had a doctoral degree in Artificial Intelligence. He was later a Signatory of the Agile Manifesto and wrote the first book on Scrum with Ken Schwaber, Agile Software Development with Scrum.
In 1970, the U.S. Air Force sent me to Stanford University to get a Master’s Degree in Statistics and I spent a third of my time for the next two years programming in LISP at the Stanford Artificial Intelligence Laboratory (SAIL). John McCarthy was the Director of the Lab and one of the founders of the discipline of artificial intelligence. He co-authored the document that coined the term “artificial intelligence” (AI), developed the Lisp programming language family, significantly influenced the design of the ALGOL programming language, popularized time-sharing, and invented garbage collection. He would come by my workstation about every other day complaining bitterly that I was using 10% of the entire compute power of the Lab on developing my smart chess program, but he never stopped me from working in the Lab.
Fast forward to 1984 and I am in Marvin Minsky’s lab at MIT as part of a Kellogg Fellowship program educating other Fellows on AI. In the corner there was a long-haired programmer. I went over and asked him what he was doing and he said he was writing a free operating system, compilers, editors, everything that was needed and launched into a long rant on how all software should be free. This was Richard Stallman, founder of the Free Software Foundation, creation of GNU, Emacs, and many free software tools, later a MacArthur Fellow and winner of many of the most prestigious awards in computer science.
By 1986, inspired by Minsky and Stallman, I had engineered my way into a CTO job down the street from the MIT AI Lab, was working with other AI researchers, co-founded Individual Inc based on the SMART linguistical analysis and documentation retrieval system from Cornell, and in 1988 joined Graphael, a LISP Object Database company. In 1989, I moved the company into a startup building to fill space vacated by Symbolics, one of the first successful LISP machine companies.
Professor Rodney Brooks and his graduate students at the MIT AI Lab came down the street to visit my company and asked if they could rent some of my lab space to start up iROBOT. Their autonomous robot, Ghengis Kahn, built on Brooks subsumption architecture, was a model of how to use simple rules to quickly evolve a smart autonomous system. Similar rules were used to create the first Scrum team in 1993.
Genghis Kahn now is in the National Air and Space Museum of the Smithsonian Institution.
When I was immersed in studying neural networks and genetic algorithms in 1987, a paper by Christopher Langdon was published out of the Santa Fe Institute mathematically demonstrating that evolution proceeds most quickly as a system is made flexible to the edge of chaos. This demonstrated that confusion and struggle was essential to emerging peak performance (of people, or software architectures, both of which are journeys though an evolutionary design space). It also showed clearly why waterfall projects slow down as you add people, methodologies, roles, meetings, and reports.
Increasing the degrees of freedom of the team to the edge of chaos and not beyond emerged from AI and is at the root of Scrum. To this day my partner, Ken Schwaber, calls his early resource web site controlchaos.com. You can read about some of this early influence at Nativity Scene: How Scrum was Born!
Because Scrum emerged from a hotbed of AI concepts, implementations, and companies, it is with great pleasure that I recommend reading this new book from one of the best Scrum trainers in the world, Paolo Sammicheli.
His keen insight into the topic and his ability to show how Scrum can produce better AI solutions brings the relationship between Scrum and AI full circle into contemporary times. It was more than 30 years ago that Christopher Langdon mathematically proved that complex adaptive systems simulated on the computer evolve more quickly as they approach the edge of chaos – the basis for team autonomy in Scrum.
Jeff Sutherland Inventor and co-author of Scrum