Industrial Artificial Intelligence
LISP
(LISt
Processor) 3 supported
representing structured programs and their
data in the same list format. Could a
computer program modify and extend itself?
Certainly, at a minimum we can accomplish
this today with discovery and tuning of
configuration parameters, and runtime code
patching is used in adaptive programming.
A UTOMATION O RIGINS
The dream of mechanical automation goes
back to the beginnings of human history 1 .
Water flow and hydraulics were used to
control movement as far back as 1500 BCE,
and steam was leveraged as an automation
force in 300 BCE. Leonardo Da Vinci designed
and implemented his Robot Knight in 1495
using pulley and cables. In the 17 th century,
automation was used to play musical
instruments and animate what appeared to
be living things. Mechanical computation to
support automation was introduced with the
Analytical Engine in 1842, and these
techniques have evolved into our modern
digital computers.
A CADEMIC O RIGINS
AI for industrial applications bridged from
academic research. Edward Feigenbaum
from Stanford collaborated with the
University of Michigan, University of
Pittsburgh and Rockefeller University on
DENDRAL 4 , an expert system for scientific
hypothesis formulation. This provided the
basis for MYCIN 5 where the techniques were
applied in medicine. Expert systems were an
attractive approach because the pruned
search of the solution space could be
explained.
Artificial Intelligence (AI) is our attempt to
replicate human cognitive function. Just like
we strived to duplicate the function and
value from animal and human labor, the
vision of machines and programs with
greater autonomy has driven research and
experimentation. Stanford University has
records 2 on AI dating back to the 1960s. Back
then a new programming language called
Neural networks, on the other hand,
stemmed from neurological science and
attempted to replicate how our brains
worked. At Cornell University, Frank
1
Spaeth, D., “From Single-Task Machines to Backflipping Robots: The Evolution of Robots”, Cutting Tool Engineering (January
2018). https://www.ctemag.com/news/articles/evolution-of-robots
2
The History of Artificial Intelligence, Stanford Libraries. https://exhibits.stanford.edu/ai
3
Berkeley, E.C., “The Programming Language LISP: An Introduction and Appraisal”, Computers and Automation (September
1964). https://archive.org/details/bitsavers_computersA_6908895/page/n15
4
Lindsay, R.K., Buchanan, B.G., Feigenbaum, E.A., Lederberg, J., “DENDRAL: a case study of the first expert system for scientific
hypothesis formulation”, Artificial Intelligence vol. 61, Elsevier (1993).
5
Buchanan, B.G., Shortliffe, E.D. “Rule-based expert systems: the MYCIN experiments of the Stanford Heuristic Programming
Project” (1984).
IIC Journal of Innovation
- 2 -