Really enjoyed this. Another layer I might add is that humans are also tuned by evolution to detect threats and opportunities. That's another relevance filter determines which signals break through and which fade into noise. It could be a helpful complement to SDT when thinking about why some alerts work and others get ignored.
Yes that’s true. I can’t remember if I included it in the article (I write them a few weeks in advance) but we are essentially change detection machines, so as long as the environment stays the same, we don’t detect any threat. It’s the change in the environment above all else which creates the loudest, clearest signal.
Thanks, Michael! This handed me language I've been missing. I've been writing about what I call "coping speed" — people moving too fast to evaluate, so they sort on cheap cues instead — and your "cutoff" is the mechanism right underneath it. Load goes up, the line goes up, and fewer things get to count.
It also left me thinking on something a bit deeper. Every enhancer you describe gets a signal noticed, but noticing something isn't the same as believing in something. A seatbelt chime is near-perfect signal design and can earn compliance, but not belief. So the strongest enhancers might be the one outside the model: a person you already trust, who both lifts the signal above the noise AND tells you what it's worth in the same move. Thank you for this — it's going to show up in something I write.
Great application. And I agree with you about belief vs compliance.
Honestly. I'm more focused on behaviors and outcomes than on beliefs. If I can put in speed bumps to prevent pedestrians getting hit by cars, I don't mind if drivers believe them to be a nuisance.
The domain of belief is best suited for aligning goals and outcomes. The means however belong to the system
Really enjoyed this. Another layer I might add is that humans are also tuned by evolution to detect threats and opportunities. That's another relevance filter determines which signals break through and which fade into noise. It could be a helpful complement to SDT when thinking about why some alerts work and others get ignored.
Yes that’s true. I can’t remember if I included it in the article (I write them a few weeks in advance) but we are essentially change detection machines, so as long as the environment stays the same, we don’t detect any threat. It’s the change in the environment above all else which creates the loudest, clearest signal.
Thanks, Michael! This handed me language I've been missing. I've been writing about what I call "coping speed" — people moving too fast to evaluate, so they sort on cheap cues instead — and your "cutoff" is the mechanism right underneath it. Load goes up, the line goes up, and fewer things get to count.
It also left me thinking on something a bit deeper. Every enhancer you describe gets a signal noticed, but noticing something isn't the same as believing in something. A seatbelt chime is near-perfect signal design and can earn compliance, but not belief. So the strongest enhancers might be the one outside the model: a person you already trust, who both lifts the signal above the noise AND tells you what it's worth in the same move. Thank you for this — it's going to show up in something I write.
Great application. And I agree with you about belief vs compliance.
Honestly. I'm more focused on behaviors and outcomes than on beliefs. If I can put in speed bumps to prevent pedestrians getting hit by cars, I don't mind if drivers believe them to be a nuisance.
The domain of belief is best suited for aligning goals and outcomes. The means however belong to the system