Doubt that Shannon's papers ever expressed anything like the analogies made in the opening of this article... This doubt infects the rest of it.
The great bulk of the articles and comments at HN are evidence of a distinction between talking about something and understanding it. Most of what's posted is what-aboutism; very little understanding.
This captures the nature and mood of the programmer today: able to manipulate interfaces, poor understanding of why things work.
The interwebz greatly expanded the realm of what-about while marginalizing and overwhelming understanding. This has been evident in news and scholarship for a generation.
The arrival of AIz are making it far worse for the reason that whereas, previously, the curious would seek those with understanding, today curiosity is being overrun by canned "answers" that cannot represent understanding because the oracle has none. Every response is a test of credulousness, with the gullible broken by hallucinations and the agents derailed by prompt injection.
The Turing test might be inverted to examine the edge at which humans can be deemed to think, given that we've already plumbed out that humans are very easily fooled by parroting.
_wire_•4mo ago
The great bulk of the articles and comments at HN are evidence of a distinction between talking about something and understanding it. Most of what's posted is what-aboutism; very little understanding.
This captures the nature and mood of the programmer today: able to manipulate interfaces, poor understanding of why things work.
The interwebz greatly expanded the realm of what-about while marginalizing and overwhelming understanding. This has been evident in news and scholarship for a generation.
The arrival of AIz are making it far worse for the reason that whereas, previously, the curious would seek those with understanding, today curiosity is being overrun by canned "answers" that cannot represent understanding because the oracle has none. Every response is a test of credulousness, with the gullible broken by hallucinations and the agents derailed by prompt injection.
The Turing test might be inverted to examine the edge at which humans can be deemed to think, given that we've already plumbed out that humans are very easily fooled by parroting.