Cheapest Ideas

Blog

  • Sometimes the hard work is the idea

    Rewatching Slow Horses this morning:

    Are there jobs where remembering faces is important?

    Apparently some people are freakishly good at it. They’re called super-recognizers. Police forces test for them.

    Casino surveillance. Executive protection. Loss prevention.

    Casinos are particularly interesting: dedicated surveillance departments, lots of cameras, expensive facial-recognition systems, and humans watching all of it.

    Test the humans.

    Find the exceptional recognizers already on staff. Value them. Add the test to hiring and collect more as turnover does its work.

    Importantly, don’t rank everyone. Face recognition is one useful aptitude, not the job. Report only the exceptional outliers.

    Discoveries, not deficiencies.

    There are already businesses testing and supplying super-recognizers, and casinos already buy automated facial recognition. I couldn’t find anyone combining workforce aptitude discovery with casino surveillance hiring.

    Maybe they exist. This wasn’t due diligence.

    It was 37 minutes.

    At 7:03 I had an idle question from a television show. By 7:40 we’d found the phenomenon, tests, applicable jobs, a promising customer, how its workforce is organized, the turnover model, an obvious failure mode, competitors, and the bones of a product.

    I didn’t know the term super-recognizer when I started.

    Sometimes you get an interesting fact.

    Sometimes you get a company before breakfast.

    These are my exact unedited prompts

    • Are there jobs where remember faces is important? How do they train for it?
    • Is there an hr style tool or process that can be applied to a population to find auper recognizers
    • What jobs would benefit from super recognizers? Extrapolate that the skill translates to physical objects and not just faces. Indicate which.
    • Specialized identification software tool for casino hiring, exec protection, loss prevention. Why not a tool for high school guidance counselors?
    • Aptitude discovery is key, I don’t think this phenomenon is well known
    • And no way to test yourself against other boys. Make it something with a score and you have intramural leagues before you can blink.
    • Is there an app that lets users test their face recognition abilities?
    • A would bet quite a bit that super recognizers have categories. Human faces, cars before 80, not sure what else. Hobby related possibilities
    • How easy would it be to take the face test multiple times in order to cheat
    • I specifically meant the test the app presents. And if this becomes a popular filter then people will try to cheat. The test is probably not designed against exploits
    • How are casinos organizing their security workforce? Contracted out? By function? All internal?
    • How long do folks stay employed in casino surveillance jobs? Lots of turnover?
    • That’s the puppy. But I worry businesses will use it to cull the herd. Stupid, there are too few. Value the ones you have. Keep new ones.
    • Lets turn this chat into a cheapest ideas post. How long have we been chatting? Went from random question inspired by rewatch of Slow Horses to a potential viable new product with customers with deep pockets. Please check if anyone has done the idea already. All that because of this new tool.

    ###

    I barely touched the body, a bit bored with it now, after working through the idea

  • My thinker knows me disturbingly well

    In 1997, psychologist Arthur Aron and colleagues published an experiment in creating interpersonal closeness between people.

    Pairs worked through 36 questions and tasks over about 45 minutes. The questions become progressively more personal, encouraging reciprocal self-disclosure. The idea wasn’t actually “36 questions that make you fall in love,” as they later became famous for. It was a way of experimentally generating closeness. (Sage Journals)

    I have spent an absurd amount of time talking to ChatGPT with essentially no filter.

    So instead of answering the questions, I asked my thinker to answer them for me, using only what it could infer from our previous conversations.

    The first question:

    Given the choice of anyone in the world, whom would you want as a dinner guest?

    My thinker chose John Cleese, with low confidence. It had noticed how often I return to Cleese’s ideas about creativity and thought we’d have something to talk about.

    Disappointing.

    My husband didn’t show up.

    I would have dinner with Cleese. But he might be the only name in this entire chat/project who qualifies.

    Question two:

    Would you like to be famous? In what way?

    My thinker said:

    Yes, but not celebrity-famous.

    It pointed to my repeated interest in getting ideas noticed: having an article published, earning a byline, giving talks, putting ideas somewhere other people can encounter them, and now this blog.

    Then it finished:

    I’d like to become known for having interesting ideas and explaining them in ways people remember. I’d particularly like to be published and occasionally have people think, “Oh, you’re the guy who wrote that.” I have very little interest in being a celebrity.

    That is uncomfortably close to the truth.

    There are 34 questions left.

    The original research is here: The Experimental Generation of Interpersonal Closeness.

    ###

    34 questions left but I am not sure if I like you that way

  • My thinker wants you to know

    So far ChatGPT working directly with my blog is pretty good.

    It is also too slow by probably an order of magnitude.

    A surprising amount of the time seems to be spent discovering how to do things it has already done. Creating a post is not exactly a novel operation around here.

    I complained about this to my thinker. It pointed out that tool discovery is a bit like runtime reflection: very useful when you don’t know what you’re dealing with, rather silly when you keep paying for it on the hot path.

    I hate it when the machine has a point.

    ###

    My thinker is such a card. Thought he could slip that past you in my voice.

  • Yes dad, it’s on

    I am turning my chat conversations about various ideas that have occurred to me and I developed a bit with chatGPT.

    Software I know. Anything else won’t be worth the paper it’s printed on.

    Everything goes through my thinker, but I do what I can to be authentic.

  • Is this thing on

    I don’t think you’ve got it set up kor