If bullshitters don’t care about the truth-value of AI, does it impose an unfair advantage?
Christopher Noessel recently wrote:
…When I type “bullshitters” here I mean it in the H.G. Frankfurt sense, i.e., someone who does not care about the truth value of what they do, only the effect. Aza Raskin has noted before that AI lowers the cost of bullshit to effectively zero, which plays into this question foundationally…
If bullshitters don’t care about the truth-value of AI, are they even capable of or susceptible to overreliance and deskilling? If they are immune, what does that mean for their use of AI? Is this a hyper-accelerated “free ride problem” or “sensible knave” problem as described by Hume? Bullshitters won’t be bound by cognitive forcing functions like the rest of us will be. Does the bullshitter’s lack of concern for truth put them at greater risk for disaster, or does it impose an unfair advantage? If it’s a short-term advantage but long-term risk, does that mean AI bullshitters just become a problem of churn?
This has special resonance for researchers when AI influencers have been aggressively pushing AI research tools for the past few years.
There were red flags:
Many of these influencers were first on the scene, building their AI brand and selling training before any rigorous validation had been conducted.
Before they became AI influencers, many had little UX reputation/credibility to speak of.
Many have since moved on to non-UX markets, coincidentally just as UX practitioners are becoming savvier about the limitations of AI.
It’s hard not to read this as bullshitting behavior.
And I think it’s clear what the consequences have been in this case:
⬆️ Short-term benefit to the AI influencers; and
⬇️ Long-term harm to how stakeholders value UX research (because of the narrative influencers pushed about prioritizing learning tools, and measuring research work by productivity metrics).
The current market for UXR is perhaps a symptom of our uncritical acceptance of this AI narrative.