Page 69 - Artificial Intelligence in the Age of Neural Networks and Brain Computing

2. Third Gen AI 57

with Stephen Wolfram, Ray Kurzweil. GAI has combined Vladimir N. Vapnik
statistical learning with knowledge rule base with reasoning system (KR). We can
also call i-AI as my-AI for machine to live with me peacefully. We identify three
missing attributes. They are (1) John Von Neumann Ergodicity principle (based
on Poincare recursive always-return-home dynamics), we replace traditional time
average with statistical 100,000 duplicated AI models. This strategy has been adop-
ted in Autonomous Vehicle (AV) by Emilio Frazzoli, CTO of Waymo, Inc. Google
applies deep learning to solve “where am I?” for the localization and mapping, using
computer vision, scene understanding (finding house number). Uber and Tesla
involve in autopilot. The progress from human-centered AV (use but don’t trust,
level 3 and 4) by 2031 (13 years) to 2050 level 5 totally automated DAVs that can
symbolize the US national flower, namely “car-nation.” DAV needs training data
from different initial, boundary, and weather conditions to save 1.3 M human folly:
drunk, drugged, distracted, drowsy driving (Fig. 3.1).
We consider a driverless car equipped sensor suite to illustrate the exemplars
(e.g., collision avoidance with all weather W-band radar or optical LIDAR and video
imaging) to 1000 identical driverless cars at the scenario of stopping at a red light.
We consider the scenario of “DAV facing the cardinal rules”: “pedestrian first”;
“red light stop”; “red light right hand turn.” We shall apply John Neumann and Poin-
care Ergodicity Theorem to consider 1000 identical DAVs equipped with identical
full collision sensor suite modeling and simulation generating an open set data called
collision FMF.
NASA began the Moon lancer, Mars landrover, DARPA began road-follower to
open field land-cruiser. Recent Science Magazine 358 (December 2017)
1370e1375, described 1 þ 5 levels of DAV by National Highway Traffic Safety
Administration (TSA) (Fig. 3.2).
It does not appear to be anywhere soon to save lives, save money, c.2070. In the
Science Magazine 358 (December 15, 2017), several important observations are
“When will we get therednot circa 2070”; “Not so fast by Jeff Mervis, 1371”;
“A matter of trust by Mat Hutson, 1375.” Following the Ergodicity Principle, we
consider 1000 to million identical systems of DAV observing the same third

FIGURE 3.1
A long history of NASA and DARPA have invested on Driverless Autonomous Vehicles
(DAV). Nevertheless, in no-man land, the scientists and technologists have not taken into
account human driver and pedestrian behaviors.