2017/2018 Edition

Three Grand Challenges

Lecturer: Prof. David Harel (Weizmann Institute).

About the lecturer | Course Summary | Relevant Material | Assignment

About the lecturer: David Harel received a BSc from Bar-Ilan University (1974), an MSc from Tel-Aviv University (1976) and a PhD from the Massachusetts Institute of Technology (1978). He held positions in IBM's Yorktown Heights research center, Carnegie-Mellon University, Cornell University, University of Edinburgh, Lucent Technologies Bell Labs and others. Currently he is a full proffessor in the Weizmann Institute of Science and heads the John von Neumann Minerva Center for the Development of Reactive Systems.
In the past he has worked in several areas of theoretical computer science, including computability theory, logics of programs, database theory, and automata theory. Over the years, his activity in these areas diminished, and he has become involved in several other areas, including software and systems engineering, object-oriented analysis and design, visual languages, layout of diagrams, modeling and analysis of biological systems, and the synthesis and communication of smell. He is an author of over 150 publications, including several widely-known general audience books like "Algorithmics: The Spirit of Computing" (1987), and "Computers Ltd.: What They Really Can't Do" (2000).
Among his awards are the ACM Karlstrom Outstanding Educator Award (1992), the Stevens Award in Software Development Methods (1996), the Israel Prize in Computer Science (2004), the ACM SIGSOFT Outstanding Research Award (2006), and honorary degrees from the University of Rennes (2005), the Open University of Israel (2006) and the University of Milano-Bicocca. He is a Fellow of the ACM (1994) and of the IEEE (1995) and was elected member of the Academia Europaea (2006).

Course summary:
I will discuss three major research directions, each of which is a "grand challenge", whose achievement (if at all) will take many years and enormous efforts. I will discuss what has to be done, what has already been done, and what the difficulties are. The three are as follows:

  1. Liberating programming from its three main constraints.
  2. Full modeling and analysis of a multi-cellular organism.
  3. A working system for odor communication and synthesis.

Relevant Material
Many papers connected with the lectures are one the prof. Harel's website. The following numbers there are particularly relevant to th three challenges:

The assignment consists of answering the four questions below.

The questions are as follows:
  1. Would we be able to remove the tractable/intractable dividing line on the sphere of algorithmic problems if someone finds a quantum n^3 algorithm for the traveling salesman problem? How about if it is shown for the generalized checkers or chess problem?
  2. Why can't we simply take the data from our 300 receptors and 40,000 detectable odorants and alleviate the need for the whiffer/sniffer cycle of work completely? How about if we are able to do this for a dog or a mouse?
  3. Would it be a good idea to have at our disposal an excellent scenario-based, inter-object development system that cannot do state-based intra-object modeling?
  4. How do you make sure that a prober in the Turing test for biology doesn't know the difference simply by asking questions he/she knows are not answerable in the laboratory?