[LINK] ANU Trusted Autonomous Systems Course 9 May via Zoom

[David]

On 3/3/23 09:24, Roger Clarke wrote:
> It doesn't seem top be offering any solutions to the conundrum, just needs, gaps, challenges and strategies, and even then only for 'communicating with stakeholders'.
>
> Oh I see, solutions are only available through consultancy assignments.
>
> Pardon my scepticism, but the emperor appears to be very lightly clad.

After our last discussion of this subject I began to wonder how one would construct a system specification and acceptance test for an AI-based system.  Maybe along the lines of "most (TBA) of the time (TBA) it will lead to acceptable (TBA) outcomes (TBA)"?

If our legal and social framework is to remain centred on humanity and the sentient creatures which inhabit this rock in the celestial void, the legal ramifications of AI are almost beyond comprehension.

But I'm sure it will lead to many consulting opportunities for as long as the buzz lasts.


On 3/3/23 10:30, Karl Auer wrote:
> "Are we all here? Good. Now listen up and listen carefully: Always remember that there is a difference between an autonomous system that IS trusted and an autonomous system that CAN BE trusted.  There are many of the former, and none of the latter. That is all. You can pick up your certificates at the door. Goodbye."
That just about sums it up!!

_David Lochrin_