Walking before we run: how biology is inspiring progress in AI

1
Walking before we runhow biology is inspiring
progress in AI

• Eldan Goldenberg Sean Psujek
• Dynamics of Adaptive Behavior Lab
• Case Western Reserve University

2
What is AI?

• Photo credits Kevin Warwick Watch (left) and
  www.irobot.com (right)

3
Good Old Fashioned AI

• TODO scan a graphic from that 1980s Time-Life
  AI book

4
Psychological background

• FIGURE a suitably flowchartish model of a task
  that is inappropriate to model in that way

5
Successes of GOFAI

• Medical expert systems
• Theorem proving
• That Paperclip (please bear with me!)
• Game players
• Photo from IBM

6
Where GOFAI fails

• Image credits SRI technology (right) and
  Stanford University (left)

7

• insert first batch of Seans slides here

8
Technology

• for this slide, do I want bullet points that
  trail the next few, or some kind of illustration?

9
Document layouts

• Figures from my MSc

10
Why not search?

• Figure from my MSc

11
Generalising the layout problem

• Photo credits University of Indiana (left)
  stonesource.com (right)

12
Mars Rovers

• Picture credits The Gadget Store, Microsoft, NASA

13
Credit scoring

• Photo credit masternewmedia.org

14
Predicting natural disasters

• Photo credits cellar.org, NASA, American-Turkish
  Association of Southern California, USDA Forestry
  Service

15
Computer Security
16

• insert second batch of Seans slides here

17
Conclusion

• Im not sure how to conclude yet, but the point
  is that this should be the last slide shown live.
  The references are just there in case we put
  this online somewhere and someone wants further
  reading

18
References Expert systems

• Exsel an agricultural expert system
  http//cnrit.tamu.edu/rsg/exsel/
• Medical expert systems in general
  http//smi-web.stanford.edu/projects/history.html
• MYCIN the most famous medical expert system
  http//www.cee.hw.ac.uk/alison/ai3notes/section2_
  5_5.html

19
References Game playing

• Deep Blue homepage http//www.research.ibm.com/de
  epblue/home/html/b.html
• A general page about AI game players
  http//www.psych.utoronto.ca/reingold/courses/ai/
  games.html

20
References Technology

• Print layouts http//www.hpl.hp.com/techreports/2
  002/HPL-2002-286.html
• Mars Rovers http//marsrovers.jpl.nasa.gov/techno
  logy/bb_software_engineering.html