Is This Anything? 24
My street has no public trash cans but each building has a domestic bin, so various people walking down our street throw their coffee cups or takeaway bags into our domestic bin. They're trying to be civic-minded and not just throw their litter in the street, and I'm sure I've done the same at times as well.
This would be fine in a place where the trash vans mechanically pick up your whole bin and dump its contents, but because we live in a city which (bizarrely) has a $14k per person municipal budget but where trash is still collected manually by hand, and where even just having bins instead of leaving out bags of garbage at the curb is a hard-won, cutting-edge innovation. Basically the trash collectors lean into our bin and grab the large bags of trash we put in there and manually throw them into the garbage truck, leaving the loose coffee cups and takeout bags behind.
This means that every week or two I have to upturn my own trash bin to pour out the loose contents into a whole new trash bag, which is a giant pain in the ass; it would honestly be easier for me if people just left their litter on my doorstep directly.
I am not sure this story is worth the time you just spent reading it, but it strikes me as a good illustration of how good intentions can lead to bad outcomes.
Many years ago, I came up with an idea for a new voting system that I called Proxy Democracy, which (in short) allowed you to proxy your vote to a friend instead of voting directly.
One noteworthy feature of this system is that it had already been invented, multiple times, by other people. But fifteen years ago it wasn't easy to find that out, because it wasn't the kind of idea that would necessarily be findable using any obvious set of terms.
I think in the old days you would talk to a knowledgable scholar in a given field and they would say "oh yes that sounds like so-and-so's work", but that's not easily scalable, and doesn't always catch repetition across disciplines.
You knew it was coming: I wonder if LLMs could solve this? And if this is just a special case of "better fuzzy search will save us a bunch of wasted effort, and we haven't really integrated that yet"?
I have a few friends who have spent/wasted significant amounts of time working on papers that it turned out someone else was already writing, and which they only found out about by (say) going to a conference 6 months into their work and meeting someone else who was clearly going to publish before they did.
This one is a little trickier because part of the problem is that the unpublished work is not yet searchable; I wonder if there's a way for people to submit their work to a registry in a way that doesn't reveal the work itself but allows other people to query and will tell them "before you write this paper, you should talk to so-and-so."
(Like many of my ideas, this is more of a social and organizational problem than a technical one – everyone would use it if everyone else did, but getting to that point is very difficult – and so a terrible startup idea unless you have the strange set of skills to get something like that off the ground).
Someone should do a psych study about how, when you get that warning that your laptop battery is down to its last 10 minutes, you can either go plug it in immediately or go "meh" and wait another 9 minutes, and then suddenly freak out and scramble to find your charger, and plug it in at the last possible second, and this behavior probably correlates meaningfully with life outcomes.