5 Comments
User's avatar
Pat's avatar

You provide a structure for understanding true intelligence and its foundations vs current AI which scrapes the surface but has no foundation.

ArnGrimR's avatar

Thank you, Pat!

That distinction between structure and surface is the starting point. The rest builds from there.

the PinkFlamingo's avatar

Statement is on point; “AI lacks accountability for errors”.

Humans should not “blindly accept” a response from AI, as a response is only as accurate as the original information provided by humans.

ArnGrimR's avatar

Absolutely correct!

Just a nuance on: "A response is only as accurate as the original information provided by humans."

That is the ceiling of the system's capabilities. It will be as accurate as the information provided by humans provided it did everything else right (including judgement when distinguishing what to use and what not use, for example).

the PinkFlamingo's avatar

Upon reading for a second time, I was struck by the recognition that intelligence and knowledge of subject matter are required to assess accuracy and take responsibility for outcomes. That said, I hope you will be addressing AI framework decisions made by lawmakers with little or no knowledge of AI itself. Thanks!!