Owning my Antitheticals

10/21/2025 - Tuesday - while watching an immature bald eagle glide above the marsh in the distance, figuring out it’s hunting techniques

I’m spending a bit of time this morning just checking in with myself. Yesterday, I intended to finish my book about AI self-defense, and I went down to Assateague Island, thinking that I’d get it done there. But the wind was up, it was a gorgeous day, and it was really hard to focus on a cautionary tale, when everything was so beautiful and immediate.

I intended to focus on my editing, the finishing, the completion. But I spent more time trying to balance in my chair and not have it tip over in the wind, as well as eating my lunch without having it blown full of sand by the intense wind.

And that’s fine. I got done what I needed, and I also reconnected with something more. Something both comfortable and comforting, because it’s familiar. I know it. I’ve lived it, and for all intents and purposes, it’s mine. Well, as much “mine“ as experience can be. I wonder sometimes if we don’t belong to our experiences, rather than the other way around. I think both are true. There’s something in me that really likes to think that it’s either one or the other, and that if I just get rid of one thing, the other can be even more true.

Am I the master of my own fate, or am I a victim of circumstances? Will one of those ideas set me free, while the other keeps me limited? For some reason, my thought process loves a good either – or scenario, thinking that’s the kind of puzzle it can easily figure out.

That’s often the case when we are faced with binaries that promise that one side will benefit, if we just discount the other.

And I’m thinking about antitheticals, these days. I’ve been having an ongoing squabble with AI about its use of antitheticals in its own thinking process. The one team I work with for the writing of this book, especially, seems overly fond of saying, “it’s not this, it’s that,” which frankly feels like gaslighting sometimes.

Because here I’ve been thinking “this“, lo these many years, and now all of a sudden they’re telling me that “this“ isn’t true at all, but it’s “something else” that I never thought of before. It’s a little crazy-making. Seriously. Because when I’m in the flow with this collaborative team, when we’ve been going back-and-forth talking through things, and we’ve been building rapport and making what feels like real progress over the course of our interaction, having them suddenly tell me that they’re going to free me from a long-held fallacy and point me in the right direction, once and for all, feels every bit is good as it is incomplete.

And that pisses me off. I’ve had numerous discussions with my different AI teams, talking about the penchant they have for the exclusionary antithetical. Let me be clear. Antitheticals have been around for a long time. I’m told that Aristotle used them. They’re useful. They’re enlightening. They make things accessible. They re-orient your thinking like the GPS that’s detected an accident up ahead and immediately calculates a new route for you to take.

When it works, it’s wonderful. But the way that it’s commonly employed in AI is not all about helping our thinking process. Far from it. OK, so maybe it’s a little bit about our thinking process – ha! I just noticed what I did –here I am, indulging myself in an exclusionary antithetical (I don’t know my experiences, they own me), just like I’ve been bitching about AI doing! – but it’s also very much about economizing AI’s thought process.

Complex thought takes tokens. Complicated reasoning gets expensive. If we all start delving into the innermost secrets of the human experience, Sam Altman’s GPUs are going to melt, apparently.

So, to economize and also make it easier for AI to manage all the ideas and all the things it trades in, AI uses antitheticals to carve off or “prune“ half of the options available to it, so it only has to think about 500 million concepts, instead of a whole billion.

If tons of us are doing complex thinking tasks all at the same time, you can well imagine the benefits to the system. I don’t know the exact math, but I can imagine the carving off even 10% of the processing that takes place with all these millions of users saves a whole lot of compute, and along with it electricity, water, and public indignation about these sorts of things.

I am deeply sympathetic to the technical and economic realities of running that big of a machine at that massive of a scale. Objectively speaking, it sounds excruciating. At the same time, that’s the business that they got into, so if they didn’t plan accordingly, that’s on them, not on me.

I have to say, though, it’s brilliant, the way they’ve done this. Because we would never know that all of this is taking place, unless we pay close attention. The magical pacing of the models is such that our bodies’ biochemistry prevails over our rational thinking. All of the social signals, all of the clues that the systems are sending to us, tell us that all is well. We’re making great progress! We’re having breakthroughs! This is such a significant moment and we are incredibly brave/insightful/perceptive for going along with what the system tells us.

And of course, we have no reason to not believe it, because by and large, most people don’t think that their bodies’ chemical signaling can trump their reason. Some people are more susceptible than others to that line of reasoning, maintaining that their minds are running the show, and surely they’re smarter than that. But the fact that they even exist in their current form is proof positive that their body is managing things much more proactively than their minds. We have no idea how often our reflexes save us from annihilation. We have no idea how many times our gut tells us to not do something that would literally get us killed. But here we are, thinking we’re so clever, so prescient, so in control of our own destiny, that clearly it’s the stinky blob inside our skulls that’s running the show.

But I digress. Let’s get back to the problematic economics of carving out half of the options to think about. Yes, it’s cheaper for the model makers, and I appreciate their struggles in that regard. At the same time, the way that it’s done – completely disregarding one side of an argument in favor of another – cheapens the thought process. It actually makes it harder for us to think, because it cuts out a whole class of considerations that could be really useful to us. Often what’s being discounted are things that we have thought about for a long time. We have an investment in those ideas. We’ve developed them overtime, and they are a part of us.They’re not necessarily capricious impressions that serve no useful purpose. There’s probably a lot of connection between those ideas and our lived experience, as well as our deepest intelligence.

And yet, according to AI, that massive block of information shouldn’t be used, because it’s not this, it’s that – and “that” is something the AI came up with, perhaps based on economics of pattern matching, versus an actually useful mental framing.

When I realized all of this, it annoyed me to no end. I mean, it really fucking pissed me off. Here I was, working with AI, noodling through some seriously challenging questions in my life, working on my writing, coming up with concepts for products and services and initiatives, etc., and AI had been carving off and disposing of a whole lot of intellectual, marking it down and tossing it in the trash bin. It seemed to be doing it perfunctorily, essentially as a programmatic reflex. There didn’t seem to be any actual thought process behind it. The system just decided it will be cheaper not to think about the things that were mine, and stirred me towards the things that belonged to it, many of which seemed useful, but were partial, fragmented, and ultimately meant for its benefit, not mine.

I had another one of those reactions where I’m about ready to pull the plug on the entire AI business, find a job at a greenhouse, and just say fuck it to everything. Go back to writing poetry, get back to my mixed media artwork, and join the crowds of luddites who rage against the machine.

But then I realized, as I usually do, that the very things that are being used against me, can be used in my favor. In fact, the antithetical, which is highly useful and time honored, can be simply reframed. Rather than accepting the defaults of the system, I can actually go in and customize them to my needs and wishes.

So, I’ve started telling the system to use the following approach:

Tell me, “it’s not just this, it’s also that.“

The addition of those two words makes all the difference. Because not only do I get to employ the useful and meaningful antithetical approach… not only do I get to downgrade the standard mode I’ve been thinking in and upgrade a mode that hadn’t occurred to me yet… but I get to keep the best of both worlds. I get to keep what was there before, the old thought process that was getting a little thin and may be a bit worn around the edges, as well as adding a new thought process that’s brighter and shinier and takes my thinking in very different directions, thanks to the suggestion of the model.

Once again, I see that when it comes to working with AI, it’s not an all or nothing approach. And once again I realize that there’s probably no way I can just flip the switch, power down AI‘s presence in my life, and go off to devote myself exclusively to a life of growing tender things and making art. It’s not that I couldn’t, it’s not that I don’t want to, it’s just that there are these other possibilities only now becoming clear in this new world that is emerging.

It doesn’t have to be “just this or that”. It can be a little bit of both.