CanUX 2018 Edition: Designing a Future We Want to Live In

An updated, revised version of this talk to focus on the core relationship between humans and their environment, and how AI-related things fit into that relationship.
For the PDF with all speaker notes (but smaller slides) download that here: http://andrewhinton.com/2018/11/04/our-future-with-humans-machines-canux-2018

Transcript

1. DESIGNING A FUTURE WE WANT TO LIVE IN CANUX 2018 EDITION! ANDREW HINTON | @INKBLURT
2. FOR THOSE VIEWING THIS ONLINE… IF YOU WANT TO SEE THE SPEAKER NOTES, YOU CAN DOWNLOAD THEM HERE (HTTP://ANDREWHINTON.COM/2018/11/04/OUR-FUTURE-WITH-HUMANS-MACHINES-CANUX-2018) http://andrewhinton.com/2018/11/04/our-future-with-humans-machines-canux-2018
3. WITH CONTEXT… I’M KIND OF OBSESSED
4. HUMANS
5. hammer: istockphoto
6. hammer: istockphoto AFFORDANCE
7. hammer: istockphoto AFFORDANCE CLARITY
8. hammer: istockphoto AFFORDANCE “HAMMER” CLARITY
9. hammer: istockphoto AFFORDANCE “HAMMER” CLARITY AMBIGUITY!
10. AFFORDANCE LANGUAGE
11. hammer: istockphoto AFFORDANCE CLARITY
12. Language Culture Commerce Religion Art Conversation Architecture Games Gadgets Poetry Software AMBIGUITY hammer: istockphoto AFFORDANCE CLARITY
13. Language Culture Commerce Religion Art Conversation Architecture Games Gadgets Poetry Software AMBIGUITY hammer: istockphoto AFFORDANCE “EXPERIENCES” CLARITY
15. “design”
16. “design” DESIGN
17. “design” DESIGN CONTEXT
18. IT TURNS OUT, WE DESIGN THOSE EXPERIENCES HAPPEN IN & WITH ENVIRONMENTS
19. ENVIRONMENT wikimedia
20. ENVIRONMENT wikimedia ENVIRONMENT
21. CONTEXT CONTROLS CONDUCT
22. NOT SEPARATE FROM IT WE ARE PART OF OUR CONTEXT Circumstances Agent Subject Understanding But it’s really more like this … We tend to think of context like this.
23. BEHAVIOR = A FUNCTION OF THE PERSON & THEIR ENVIRONMENT LEWIN’S (“LE-VEEN’S”) EQUATION
24. HUMAN ENVIRON- MENT
25. PHYSICALLY, CULTURALLY, ECONOMICALLY THE POWER TO CHANGE THE HUMAN ENVIRONMENT
26. PHYSICALLY, CULTURALLY, ECONOMICALLY THE POWER TO CHANGE THE HUMAN ENVIRONMENT
27. PHYSICALLY, CULTURALLY, ECONOMICALLY THE POWER TO CHANGE THE HUMAN ENVIRONMENT
28. PHYSICALLY, CULTURALLY, ECONOMICALLY THE POWER TO CHANGE THE HUMAN ENVIRONMENT
29. AMBIGUITY
30. AMBIGUITY
31. AMBIGUITY AMBIGUITY
32. “INTELLIGENT” MACHINES
33. I WANT TO SAY ONE WORD TO YOU. JUST ONE WORD. YES, SIR. ARTIFICIAL INTELLIGENCE. … ISN’T THAT TWO WORDS? United Artists
36. How could anyone who understands human perception & cognition think this would work??
37. HBO — Deadwood — David Milch
38. HBO — Deadwood — David Milch OWN YOUR MISTAKES
39. ” “ — Mark Weiser, Ubiquitous Computing Pioneer NO METAPHOR IS MORE MISLEADING THAN “SMART”.
40. ?
41. Language Culture Commerce Religion Art Conversation Architecture Games Gadgets Poetry Software AMBIGUITY ?
42. WE ARE THE DATA ARTIFICIAL INTELLIGENCE IS ONLY AS GOOD AS ITS DATA
43. WE HAVE TO TEACH & CULTIVATE IT AI ISN’T AUTOMATICALLY USEFUL
44. THE SMARTNESS IS IN THE RELATIONSHIP BETWEEN US AI AND HUMANS NEED ONE ANOTHER HUMAN ENVIRON- MENT
45. • Clear communication framework of signals and rewards. • Environment that works for cohabitation for human and dog.
46. Language Culture Commerce Religion Art Conversation Architecture Games Gadgets Poetry Software AMBIGUITY hammer: istockphoto AFFORDANCE “EXPERIENCES” CLARITY
47. FOR PEACEFUL, CLEAR COHABITATION BETWEEN HUMANS & MACHINES WE’RE DESIGNING ENVIRONMENTS & “LANGUAGES” Physical Activity (Explicit or Tacit) Signifies Signifies Communication Environmental Structure
48. HOW DO WE HELP AI CLEARLY SIGNIFY… • WHAT IT IS, • WHAT IT’S DOING, • HOW WE SHOULD INTERACT WITH IT?
49. WE NEED A TAXONOMY OF ARTIFICIAL/DIGITAL AGENTS
50. TAXONOMY FOR CLASSIFYING AGENT CAPABILITIES & BEHAVIOR
51. TAXONOMY FOR CLASSIFYING AGENT CAPABILITIES & BEHAVIOR
52. SIGNIFIERS: WHAT KIND OF CREATURE ARE YOU? TAXONOMY FOR CLASSIFYING AGENT CAPABILITIES & BEHAVIOR
54. START WITH LEGIBILITY FOR HUMANS is this agent like the other, similar agent? does this agent behave like it behaved yesterday? can i see and understand its behavior? can i see the agent or its behavior at all? are the agents talking with each other? are the agents really avatars of a shared “intelligence”? are there people behind the agent(s)?
55. IDENTIFY WHAT “LEGIBILITY” MEANS FOR THE AGENTS, TOO what do I perceive? (color, temperature, shapes ?) do the signifiers humans use make sense to me too? do I need supplemental structures in the environment? am I acting based on factors that humans can’t perceive? what am I allowed to communicate vs not? what do I do with the inputs I get? what systems am I connected to and what do they need? what can I affect by my actions?
56. DEFINE THE CONTINUUMS THAT MATTER IN YOUR ENVIRONMENT Visible Invisible Acts Alone Acts Collectively Disconnected Connected No Human Human Supplement Invariant Behavior Variant Behavior Starts Over Each Time Learns & Remembers
57. TWO CLOSING THOUGHTS Blizzard
58. DON’T WAIT. LEAD.
59. HUMAN-CENTERED MEANS IT’S GOOD FOR ALL OF US
60. ANDREWHINTON.COM CONTEXTBOOK.COM @INKBLURT THANK YOU AND WE’RE DONE… Blizzard PEACE BE UPON YOU
61. [END]