AI BS

Brian Dunaway wrote:

I have a number of opinions regarding AI – nearly all negative.

But, for one, AI is supposed to be that by which humankind accelerates toward the unfathomable event horizon of understanding, the exquisite singularity, touching the very face of God!

Uh-huh.

Except that every. single. day I receive an erroneous answer from some AI algorithm. Every day. Not that I ask that many questions – in fact, rather infrequently, but I would say at least a THIRD of the time I get a WRONG answer. And I don’t mean a biased emphasis or some alternate way of looking at something. I mean just WRONG.

In the example below, there is no subtlety. It is simply WRONG. Hexadecimal employs characters 0123456789ABCDEF. But instead of this base 16 system, this example appears to suggest a ((2×26)+10) base 62 system. Just WRONG.

Not all AI is equal, to be certain. For example, I have found far fewer errors in Google’s AI than Bing’s. I always compare for grins.

Thus far, that event horizon looks more like the swirling water in a toilet bowl. I can think of other more entertaining ways to collapse the entire northeastern power grid.

The Best of Llewellyn H. Rockwell, Jr.