Original Source Here
Allen Newell: Intelligence Engineer
I first encountered Allen Newell in 1962 at Carnegie Tech (now CMU) in a graduate seminar . One day he asked a graduate student, David Parnas (who went on to an influential career in software engineering) if he would make a presentation the following week. Parnas said, “Well, if you twist my arm….”, so Newell, always open to a joke, walked over to him and twisted his arm until he submitted.
In that seminar, and many times thereafter, I witnessed Newell’s intellectual fearlessness. His lifelong research goal was understanding human cognition, and he was a founder of a the field of cognitive psychology. But he would walk into any academic area and engage the experts in intellectual exchange. In one period he partnered with Gordon Bell (a founder of Digital Equipment and architect of the PDP series of computers) and Dan Siewiorek to create a descriptive theory of hardware architecture. When a student claimed in his Ph.D. oral that the system he had designed was too hard to program, Newell said “George (Robertson) and I could do it in a weekend.” and proceeded to spend the next 48 hours proving it.
He embodied the concept of interdisciplinary research blending social sciences with technology. Carnegie Mellon now boasts of it as part of its brand but often fails to perform it; it takes confidence and courage to play with another department.
Later he spent many months at Xerox PARC working with two of his Ph.D. students, Stuart Card and Tom Moran. Their work led to the field of Human-Computer Interaction.
He used his intellectual fearlessness and general good nature to guide many phases of Carnegie Mellon’s evolution. In the early 1980’s he led a campus-wide initiative to create a modern campus wide computer system based on personal computers and networks. The first step was convincing the administration and key departments that it was a good idea. He supported President Cyert’s effort to get IBM to pay for the whole thing. He was persuasive in convincing me to leave Xerox PARC to direct the development of what became the Andrew system. In one discussion about my resistance to working for the much-feared IBM he shocked me by saying, “You’ll find that violating taboos can be quite enjoyable.” He counseled me often about the system’s goal, saying “The network is more important than the personal computer” — something that has certainly proven to be true.
He was a key persuader in the creation of the School of Computer Science. Any veteran of intercollegiate politics can appreciate the skill that was required. His argument was that every major move that the burgeoning Computer Science ended up being decided in the President’s office, and any intermediate administrators, no matter how astute, would be superfluous.
Late in the 1980’s, he called an ad hoc campus meeting to initiate a departmental-level effort in Human-Computer interaction. There was an overfull room and great enthusiasm, but nothing happened. The main departmental players, Computer Science and Psychology were crystallizing (not to say ossifying) into disciplines that had other focuses. It was later left to me, his student Bonnie John, and Dean Raj Reddy to create the current HCI department.
Newell and Simon are credited with identifying list processing as an important software technique and they invented a programming language, IPL (Information Processing Language), based on list processing. The programmed many of their famous systems, e.g. The Logic Theorist, in it. Other than supporting lists as basic structures, it was a horrible language combining slow execution, the syntax of machine language, without even the mnemonic names of assembly language. It was quickly displaced by John McCarthy’s brilliant LISP language. Newell was once asked why IPL lacked mnemonic names and replied “In my program K298 is totally mnemonic to me. It stands for the concept of ….”
Allen had a wonderful attitude about research and the inevitable conflict of ideas. There are often starkly differing opinions about how to do things. Whenever he encountered such conflicts, often held by different researchers, he’s say “Let us devise an experiment to understand the tradeoffs.” Furthermore, he was entirely willing to be proven wrong. For example, given his interest in human-cognition he believed that AI programs should be based on simulating human thought. As time went by, it became apparent that the exponential improvement in hardware speeds allowed computers to solve problems like playing chess through brute force. He freely admitted defeat in his engineering approach while continuing his research into human thought.
Newell and his senior colleagues seemed to have little interest in undergraduate education. The 1980 computer science department looked more like the Rand Corporation with teaching pursued as a hobby. It fell to the founders successors, Nico Habermann, Allen Fisher, and others to change course.
In the summer of 1999 Allen was dying of cancer. I broke out of business relationship to say “You know, I am really sorry this is happening.” Allen said “Not as sorry as I am!” His subsequent death nudged me into returning to Carnegie Mellon from a leave and becoming the head of the CS department because I felt the institution he created might falter. (Little did I know that Nico Habermann, a very important leader would die a few months later.) My first impulse was to try to perpetuate Allen’s attitudes to all the faculty as we mourned his passing. Unlike Simon, he never wrote an autobiography; and, unlike Perlis, never left trail of epigrams. I decided to make up a list of precepts which sometimes appeared on the doors of faculty offices. Here they are:
- Do what you love, love what you do.
His incredible energy and enthusiasm sprung from this.
- Help others to find a similar state, no matter how different their choices might be.
Since he was happy and secure in the rightness of what he was doing, Allen was open-minded about what other people did, and often could help them make good choices.
- Don’t worry about how intrinsically smart you are or anyone else is.
I never saw him feel threatened by another person’s brilliance or offended by their lack of it. He only judged performance.
- Be intellectually tough — uniformly on everyone.
Allen was intellectually the toughest critic I ever had. At first, I didn’t think this fit with his supportive attitude. But then I realized that it wasn’t personal; he applied tough standards to himself and everyone else.
- Be careful about what you commit to do, and then really do it.
He agreed to do only some of the things I asked him, but he would always help with something he believed was important. When he did, there was no doubt about his level of effort.
Whatever his other obligations, Allen was always present for the department’s Black Friday meetings in which the entire faculty reviewed the progress and prospects for each Ph.D. students. He used these occasions to train all of us on how to evaluate research. I remember one case, I think of his own advisee who was generally respected and was in his ninth year of study. It was time for him to graduate, but Allen refused to let him go because he had yet to empirically prove the hypothesis of his thesis.
More about Newell can be found in my memoir What Were We Thinking?: Reflections on My Sixty Years in Computer Science, to appear on Amazon KDP in the fall of 2021.
Trending AI/ML Article Identified & Digested via Granola by Ramsey Elbasheer; a Machine-Driven RSS Bot