Hi,
I have a pleasure to share my full conversation with Professor Missy Cummings, the Director of Duke’s Humans and Autonomy Lab, and one of the first female fighter pilots in the US.
Missy and I discuss:
We are not even close to developing fully autonomous cars, because (i) autonomous systems cannot derive abstract learning principles like humans, and (ii) although we know how to test autonomous systems that use deterministic code, neither academic nor industrial experts know how to stress test the probabilistic code used in most driverless cars.
This striving for a new level of autonomy and innovation is critical for moving the football of development down the field - but needs to be tempered with technorealism.
The move upwards across levels of automation is not linear. Whilst early developments may take similar amounts of time and money, the final advancements demand exponentially more.
‘More data’ will not resolve the gap in abstract learning in autonomous systems; it is not yet even…
Keep reading with a 7-day free trial
Subscribe to Exponential View to keep reading this post and get 7 days of free access to the full post archives.