Between a rock and a hard place: Cognitive Science principles meet AI-hard
Artificial Intelligence and Cognitive Science have always
been overlapping disciplines.
Early in their history, that overlap was considerable. Herbert A. Simon wrote that ³AI can
have two purposes. One is to use
the power of computers to augment human thinking. The other is to use a computer¹s artificial intelligence to
understand how humans think.²
Conversely, at the foundation of the Cognitive Science Society, Artificial
Intelligence was identified as one of the core constituent disciplines. However, over time the two disciplines
have increasingly diverged under seemingly incompatible constraints. The difficulty of many problems tackled
by AI led it to adopt brute-force or domain-specific solutions that arguably
were not cognitively plausible.
Conversely, the need for precision and reproducibility increasingly led
cognitive science to focus on experimental paradigms that AI did not recognize
as hard problems. Recently, in his
AAAI presidential address, Tom Mitchell called for a rapprochement between the
two disciplines on the basis of convergent evolution.
In this symposium, we will explore what each discipline can
expect to contribute to the other in the relatively near future. More specifically, we will ask whether
cognitive science approaches can be applicable to AI-hard problems. Conversely, what impact can the challenges
of AI-hard problems (and the techniques that have been applied to solve them)
have on cognitive science? To
improve focus and enliven debate, we will structure the symposium around a
number of questions with clear answers outlining sharply divergent positions. These questions and the resulting dual
Is cognitive science relevant to AI problems?
human mind is one of the most flexible, general and powerful thinking
devices possible. The right
way to solve AI problems is the cognitive way.
No. Many of the workings of the human mind are
specific to its substrate and problems. Cognitive methods have failed before and will fail
Are ³good enough² solutions valuable?
Too much focus has been devoted to the search for optimal
over a broad range of problems and computational tractability are
essential to long-range success in cognitive computing.
Optimal performance is essential, even if it results in high
Moore¹s law and distributed computing will provide the scale
needed to reach brain-like computing power within a generation.
Are multi-level heterogeneous approaches
- A. Yes. Most problems are not
amenable to a single approach but instead demonstrate aspects best
addressed at different levels, such as numerical techniques,
expertise-based pattern matching or meta-level reasoning.
- B. No. Problems are typically
best approached using a single, targeted method (e.g., search in
chess) with maximal efficiency. These methods can be hand-selected by human
- 4. Is
adaptiveness an essential component of intelligence?
- A. Yes. A key component of
human intelligence is its ability to adapt to constantly
Hardwired solutions are brittle and in constant danger
of becoming obsolete.
- B. No. Adaptiveness is often
a result of insufficient information or ill-defined
problems. When a
problem is well defined and understood, the best solution can
be found and developed once and for all.
- 5. Are
the most efficient solutions problem-specific?
- A. Yes. Problem-specific
solutions will always be the most efficient because they do
not have to include the overhead of generality.
- B. No. Real-world problems
are so complex that they include many different aspects
which require significant generality to be solved in a
Instantiation of general knowledge and capabilities
and re-use of capabilities across problems are two possible
mechanisms to provide the needed cross-problem generality.
- Is developing specialized, modular
components a reasonable way to study general intelligence?
- Yes. The significant progress in both AI and
cognitive science have demonstrated that studying
individual phenomena largely in isolation is necessary for
scientific progress in the field. Recent
experimental findings in neuroscience also support
modularity as a core principle of neural organization.
- No. A core feature of intelligence is the
ability to bring disparate information together. Modular approaches
to cognition and intelligence will not scale because assumptions
within modules are incompatible across the modules.
- Can Artificial Intelligence contribute
to our understanding of human cognition?
- A. Yes. Artificial
Intelligence addresses many problems central to human
cognition in an effective, functional manner. AI solutions
reflect fundamental constraints that are applicable to all
approaches to those problems, including human cognition.
- B. No. Artificial
Intelligence provides an ad hoc approach to problems that
sacrifices generality and plausibility for the sake of
Cognitive solutions to problems are fundamentally
different from AI solutions.
will be asked to take positions on one or more of these
questions by using examples from their own work or
synthesizing other works into coherent patterns. Contributions can be
in the form of either full-size 6-page papers or short 2-page
white papers. The
workshop will be structured in a number of sessions each
centered around one of these questions. Each session will include
two to four presentations arguing both sides of the issue,
followed by a brief closing argument and a substantial
discussion session between panel presenters and the workshop
The list of
questions should not be considered a definitive list. Indeed, we would
welcome the identification of additional issues that bring up
new connections between the Cognitive Science and Artificial
Intelligence communities. However, Contributors who focus their
submissions primarily on recent progress in their specific
areas of research, unless it is directly and obviously
relevant to the general goals of the workshop, will be