
I have a problem with ideas. By the time I notice one, it’s already there.
That makes investigating the damned things inconvenient. One moment I don’t have the idea. A moment later I do, and now I’m standing there admiring the finished product when what I really want to examine is whatever happened immediately before it arrived.
People considerably smarter than I am have been worrying about where ideas come from for a very long time, and some of them have produced answers containing words I would have to look up before I could disagree with them. I’m approaching the question from the other direction. I’ve had ideas all my life, some useful, some not particularly useful, and a few that probably should have been escorted quietly from the premises. I’m trying to catch one coming through the door.
Suppose an idea occurs to me while I’m sitting here. If the finished idea had already been sitting somewhere in my head waiting to be discovered, finding it sounds suspiciously like remembering rather than inventing. But something was certainly there before it appeared. I had memories, experiences, things I’d seen and heard, people I’d known, mistakes I’d made, things I’d built, things I’d broken, things I’d read, places I’d been, and an impressive collection of miscellaneous mental debris whose continued storage could not always be justified.
Maybe the idea wasn’t there yet. Maybe the pieces were.
Memory May Be More Than Remembering
We usually talk about memory as though its principal job is to help us recover the past. Where did I put my keys? What was her name? When did we buy that house? What did the doctor tell me? Those are certainly useful services, particularly the keys, but I’m beginning to wonder whether memory has another importance that we don’t notice because we’re too busy trying to remember where we left the damned things.
Memory gives us something to think with. When I encounter something today, I don’t meet it with an empty head. Whatever I see, hear, feel, smell, read or otherwise encounter arrives in a mind already carrying the accumulated remains of earlier encounters. Most of that material isn’t consciously lined up waiting for inspection. I couldn’t give you an inventory of it, and some of it I probably couldn’t deliberately retrieve if you paid me. Yet something happening now can suddenly make an old piece available.
That happened in the New Mexico story I told elsewhere on this site. I woke from a doze in a car, looked across the desert at a plateau, and immediately populated the ridge with imaginary Indians from the Westerns that had filled my younger years. Nobody asked me to do that. I wasn’t solving a problem. Something I saw outside the car simply bumped into something already in my head, and for a few seconds the two belonged together.
Civilization advanced very little as a result. But the event interests me now because I wonder whether something similar happens when the result is useful. Perhaps memories don’t contain our ideas. Perhaps they provide some of the pieces from which ideas can emerge.
I Think I See Mine Before I Can Say Them
There is another part of this that I don’t know how to prove, so I won’t pretend I can. I can only tell you what it seems like from inside my own head. When an idea first appears, I often seem to see something before I can explain it.
I don’t necessarily mean a photograph-like picture. It may be an arrangement, a movement, a relationship, a shape, some numbers, part of a machine, or something closer to a dream. Sometimes the whole thing seems to be there for an instant, but when I try to put it into words I discover that language is considerably slower than whatever just happened. On occasion the damned thing can even disappear while I’m trying to describe it.
Perhaps your thoughts don’t work that way at all. Maybe your ideas arrive as words. Maybe you hear them. Maybe they’re mathematical. Maybe you feel a relationship before you can see or describe it. Maybe the form changes depending on what you’re thinking about. I don’t know, which is precisely why other people’s accounts become useful.
Whatever that first internal thing is, though, sooner or later we face another problem if we want anybody else to know about it: we have to get it out of one head and somehow cause something useful to happen in another.
Build Me a Little Red House
Imagine a little red house beside a river.
You probably have something now. So do I. But I doubt very much that your little red house is mine. Mine has already acquired characteristics I never gave you. I can see its position relative to the river. There are trees nearby. The river has a particular width. I have some notion of the house’s proportions and where I’m standing while looking at it.
Your river may be muddy and broad while mine is narrow. Your house may have white trim. Mine may have a porch because apparently my internal planning department approved one without consulting me. Yet despite constructing two different places, you and I can have a perfectly sensible conversation about the little red house beside the river.
I find that remarkable. I didn’t put my little red house into your head. I gave you a few words. Those words encountered whatever you already had available, and you constructed your own house well enough that we can proceed as though we’re discussing the same one.
Most of the time, that’s good enough.
But Good Enough May Be Where Ideas Get Interesting
Suppose I am trying to explain an idea instead of a little red house. I talk, draw pictures, make a model, show you numbers, wave my hands around or resort to whatever else becomes necessary when you continue looking puzzled. You use all of that to construct your understanding of what I’m trying to show you.
But you aren’t constructing it with my lifetime. You’re using yours.
Something I consider unimportant may remind you of an experience I’ve never had. You may misunderstand part of what I meant and accidentally notice a useful relationship. I may describe a problem from one business and you recognize something similar from an entirely different occupation. You may ask a question that seems peculiar to me because I don’t possess the experience that made the question obvious to you.
Then you tell me what occurred to you. Now I have to reconstruct that using my own collection again, and your addition may bump into something else neither of us had been considering.
Well, I’ll be damned. Maybe we aren’t simply passing finished ideas back and forth. Maybe we’re repeatedly putting incomplete representations of them into different collections of experience and seeing what each collection does with them.
That begins to explain something I’ve observed throughout my life. An idea can become more interesting when another person gets hold of it. Not always, of course. Occasionally another person improves an idea enormously by killing it before considerable money is lost. That too is a valuable contribution. But sometimes another person sees something the originator couldn’t see because the other person has something different to see with.
And Then Along Came Artificial Intelligence
This is where I need to be careful, because I’m currently doing a substantial amount of thinking and writing with artificial intelligence. AI can produce combinations, suggestions, questions and apparent ideas at an astonishing rate, and I use it for exactly those reasons. But I am not prepared to say that whatever happens inside an AI system is the same thing that happens when an idea occurs inside a human mind.
For these stories, I’m going to keep the distinction visible. A human has spent a lifetime encountering the world through an extraordinary assortment of senses and bodily experiences. We see and hear, certainly, but we also experience pressure, temperature, balance, position, pain, internal bodily states and other signals, many of which we barely notice until something goes wrong. Some encounters repeat. Some have consequences. Some become extraordinarily important. Others apparently disappear until something unexpectedly brings them back.
AI receives and processes information by a very different route. Whether increasingly capable AI systems will eventually do something we should call independent ideation is a question I’m quite happy to leave open. I don’t know, and pretending otherwise would spoil a perfectly good mystery.
For now, the distinction that interests me is simpler. I have a lifetime of encounters that happened to me. AI has learned patterns from enormous amounts of human-created information. Those are both remarkable resources, but they are not obviously the same kind of resource. Working with AI gives me another place to watch combinations occur, another source of questions, and occasionally another participant capable of knocking my thinking sideways.
I want to know what difference that makes.
So Did Memory Cause the Idea?
I started this line of thinking with a fairly simple proposition: memories cause ideas. I don’t think I would say it quite that way now. Cause is too strong. A warehouse full of memories doesn’t automatically manufacture innovation any more than a garage full of tools automatically builds a chair.
What I’m circling now is more tentative and, to me, considerably more interesting. Memory may provide some of the inventory with which thought proceeds. A new encounter makes some portion of that inventory relevant. Often an existing answer is perfectly adequate and the matter ends there. Need satisfied. Nothing new required.
But sometimes the familiar answer doesn’t quite fit. Something is missing, unsatisfactory or simply odd enough to keep our attention. Examination continues. Other memories, patterns and pieces become relevant, perhaps including something from a completely different part of life. Occasionally pieces that had never belonged together before enter a relationship that works.
Maybe somewhere in there is what we eventually call an idea.
I don’t know yet, and I rather like not knowing. It gives us something worth looking at instead of something we’re obliged to defend.
Can You Catch One of Yours Arriving?
My own memories can tell me only what thinking seems like inside one peculiar human mind, and I’ve been living with that fellow long enough to question the reliability of the witness. What I need now are other witnesses.
Think of an idea you’ve had—not necessarily a grand invention, just something that genuinely occurred to you. See whether you can get close to the moment before it appeared. Were you looking at something? Talking to somebody? Trying to solve a problem? Did an old memory suddenly become relevant? And when the idea first appeared, what was it like? Words? A picture? Movement? Numbers? A relationship? Something you still can’t describe properly?
Then see whether you can trace any of its pieces backward. Had you encountered them somewhere else before? Did something from years earlier suddenly become useful? Did another person say something that caused you to reconstruct the problem differently? Perhaps you can’t trace it at all. That answer is useful too.
If you can get anywhere near that moment, bring the experience over to Your Ideas and tell us what happened before the idea became an idea. Don’t clean it up into a brilliant finished invention. I’m increasingly interested in the messy bit—the half-formed picture, stray memory, conversation, wrong turn or peculiar connection that came first.
Because by the time the finished idea announces itself, we may already have missed the most interesting part. I want to catch one a little earlier and ask it the question that started this whole damned investigation: Where the hell were you before I thought of you?
