Our AI-Powered Post-Truth, Post-Trust Unraveling

You are receiving this post/email because you are a subscriber to Of Two Minds / Charles Hugh Smith. This is Musings Report 2023-43.

Self-proclaimed techno-optimists giddily assure us that AI (artificial intelligence) is the salvation of humanity because AI "solves problems." As long-time readers know, I've been following AI for 40 years and large-language model (LLM) AI doesn't seem all that "intelligent;" as I outlined in Musings 23 (Will AI Deliver the Worst of Both Worlds? ),

LLMs have replaced previous computer-human interfaces with natural language facility. This makes it very easy to over-estimate the actual "intelligence" behind the interface.

More robust AI may deliver protein-folding solutions, but it can't explain how it arrived at the solution; it's a "black box." If AI is unable to explain what it's doing, is this actually intelligence? 

What the techno-optimists fail to mention is that AI doesn't just "solve problems," it creates problems, too. The techno-optimists have a ready solution to any AI generated problems: more AI. Which then generates a new set of problems, which we then attempt to remedy with more AI, until the AI "solution" consumes itself.

You've probably seen the phrase Post-Truth, generally defined as "objective facts are less influential in shaping public opinion than appeals to emotion and personal belief."

As an independent on the fringes of public media (oftwominds is bit of rock out in the Oort Cloud), I would say Post-Truth also means "the suppression of evidence that is counter to the approved narrative." 

Those who are not in public media--bloggers, video-bloggers, podcasters, etc.--are generally unaware of how many independent voices have been banned or shadow-banned for even hinting that the approved narrative isn't the sum total of what's true.

For example, a podcaster recently told me that he was banned by a major social media platform for speaking with an academic scientist about vaccine approval protocols. A financial blogger's wife who grew up in fascist Portugal as a child, told me he was also shadow-banned for something a guest said. Needless to say, she reported that it reminded her of her childhood.  She said they couldn't risk having me on the program again as a guest, so sorry.

My own experience of being shadow-banned started back in 2016, winning me a coveted spot on the infamously false PropOrNot black list published by the Washington Post. The tentacles of this black list are long; I've had readers report all manner of evidence of shadow-banning.  Once you're on a black list, there's no procedure for getting off it.

I recently started reading Franz Kafka's last novel, The Castle. It's broadened my understanding of Kafkaesque. The entire phenomenon of being banned / shadow-banned is very reminiscent of Kafka's work, as your "crime" is not specified, you do not get a trial to contest your conviction, and there is no process to remedy a wrongful conviction (an appeal).

Kafka's The Castle adds layers of nuance to Kafkaesque. What people know is obscure, how much authority they wield is obscure, the procedure to gain access to the Castle is obscure, what they have accepted as true is obscure.

Thanks to AI tools that make it easy to create "deepfake" videos and audios, Post-Truth is now becoming Post-Trust: since everything can be faked, nothing can be trusted.