2018.03.27 Meeting Notes

INCOSE Augmented Intelligence Challenge Team

2018.03.27 Meeting Notes

Attendance: [X] Mark Petrotta, [X]Troy Peterson, [X]Bill Schindel, [X]Jon Wade, []Jimmy McEver, [X]Donna Rhodes

Recap of last meeting

Types of learning:

INLINE – Real time

AFTER THE FACT – what happened?

Chess AI -> learns YOU -> helps you

Decision -> action -> AI analyze results

Person <-> AI

               Helps you learn

Learning & Knowledge Manager vs LC Manager(gears)

Learning & Knowledge Manager	LC Manager(gears)
Learn new things	Not learning new things
Reflecting on past history	Exploit what is already known Apply what is already known
Applies to ISO15288 processes	Applies to ISO15288 processes

AI Waves:

1)      80’s AI: Knowledge capture

2)      ML/Pattern matching

3)      Contextual information

Advertising – personalization

Augmentor vs Augmentee

               What if both human?

               Some people not augmentable??

              

Studies show how people remember where something is better than fact itself.

Learning connected to retrieval

Understanding certainty (uncertainty) of AuI

Inject noise -> diff result?

Sensitivity analysis

HIL (human in loop) -> avoid Armageddon

Competition vs Cooperation

Two Major Forces in Universe: 

1)      Entropy  - increasing

2)      Self-Assembly Complexity

Taxonomy – how we can think about AuInt

               How impact design / engineering of systems?

               On-going vs algorithmic (distinction?)

TO DO:

1)      Put notes on wiki

2)      Add team to wiki?

3)      15288 architecture in Donna’s slide

4)      Design patterns