Human Fallacies
"Errare humanum est, in errore perseverare stultum."
It is in the human nature to be wrong sometimes or even oftentimes. We know this. We found ways to deal with it. We know that holding on to a false belief, a disproven idea, or an unsatisfiable hope only leads to disappointment.
Yet, it is human to fall for the same logical traps (aptly call fallacies) and cognitive biases over and over again. It happens. There is no shame in it. But insisting on the error once it has become apparent is stupid.
This is known since ancient times and hasn't changed since.
The Freak Wave
Our fallacies sometimes add up, hit at the same time, and reinforce each other with increased impact - like waves in a storm that are manageable one at a time but when peaks reach us in unison, they can produce a devastating freak wave.
AI is such a freak wave.
It hits many of our weak spots at the same time. Here is a small excerpt of the very long list:
- "If it can do this then it can do everything." (undersampling)
- "It understands me." (Eliza effect)
- "It produced what I needed." (Barnum effect)
- "It is intelligent." (anthropomorphize)
- "It will make me rich." (greed)
- "Everybody is doing it." (perceived mainstream)
- "It's easy and convenient." (laziness)
- "I will be left behind." (fear of missing out)
- "It will solve my most pressing issue." (hope for salvation)
The full list is so long that it will most likely contain your most pressing issue or at least enough issues that resonate with you.
It is more devious than a simple "snake oil" dealer that only hooks on to one false hope. It is more capable than ordinary con artists because of the absurd amount of money behind it. It is more devastating than a single seductive leader because it reaches a whole society by reaching every single one personally.
Riding the Storm
With so many storm clouds on the horizon, the outlook might appear bleak. But with a better understanding of our fallacies we might be able to better recognize their exploitation and prevent damage.
To this end, I plan to present these fallacies and their application in the context of AI in a series of documents. I'm happy to receive suggestions via GitHub issues.
Issues that are AI generated will be flagged as such and otherwise ignored. Since the AI crowd is notorious for silencing critics by flooding open conversations with DoS attacks, I might be forced to close down the service in the future. I apologize in advance.
Stay tuned for further content.