You probably left last week satisfied with yourself.
Not about the decision — that is still on the table, exactly where it was. Satisfied with yourself. Because last week you did something clever, and you knew it. You caught yourself. You had been carrying a theory of wisdom your whole life without ever examining it, and in an hour you stated it out loud, chose your answer — the store, or the act, the thing you have that the machine hasn't — and then you noticed a second claim that had come with it unannounced. And I've got it. You saw it arrive. You caught the flattery that had slipped in behind something that looked like reasoning. And catching it felt good. It felt like exactly the sort of thing a person with good judgment would notice.
Tonight we find out whether that satisfaction was earned.
Here is what I did not tell you last week. Catching the second claim is not the hard part. The hard part comes when you stop merely looking at your theory of wisdom and ask it to do its job. Last week we made it visible. We held it up and turned it over. But we never asked it to work — never set it down and said: now do the thing you have quietly been promising to do all along. Rule the machine out, and tell me I have it. Both, at once, out loud.
That is tonight. And I will say it now, so it does not take you by surprise: the satisfaction you left with last week does not survive the test. Not because I am going to tell you that you are wrong — I am not. Because your own theory, the one you committed to, comes apart when you finally make it carry both claims in the open. And it comes apart in your hands, not mine. I am only going to hand your answer back and ask you to run it. You will do the rest.
This is Philosophy for Us — philosophy for everyone, no degree required.
This is the sixth of eight episodes on what wisdom is, so let me bring us up to the edge of it. Last week the show stopped surveying. For four weeks I had set out five accounts of wisdom, one after another, other people's, for you to consider from a safe distance — and then last week I put them down and asked the question I had been circling: which one is yours? What is the thing you believe you have that the machine hasn't? And you answered. Most of you landed in one of two ordinary shapes: the store — everything you have lived through that the machine has only read about; or the prouder one, the act — the judgment, the call that you are the one who has to make. You chose one. You committed. And then you noticed the second claim you had folded in beside it without seeing it: that whatever wisdom is, you have some, and it is yours. One sentence that was really two — the machine isn't wise, and I am. You had carried it as a single fact your whole life. You left having established only one of the two claims.
Tonight we go after the other one. We take the answer you committed to — the store or the act, in your own words — and we make it do both jobs at once, and we find out what that costs.
So go and get it: the answer you committed to last week. Whichever one you would defend to the end — the store or the act, the years you have lived or the call you make — pick it up and hold it where you can see it, because everything tonight happens to that answer, and it only works if it is honestly yours.
Before we run it, let me say plainly what your answer has been promising you, because you have never made it say the promise out loud. Your theory of wisdom, whatever you chose, has been doing two jobs for you all this time, and doing them so smoothly that you took them for one. The first job: it marks off something and says the machine is not part of it. That is the job you hired it for — the one you have been grateful for, the one that lets you sleep a little easier in a year when a machine on your desk knows more than you ever will. The second job is quieter, and you have not been watching it: it marks off that same thing with you counted inside. It does not only say the machine lacks wisdom. It says you have it. That second job is the one you never checked. That is the second claim from last week. Tonight we stop letting it pass unexamined. Tonight we ask your theory to do both jobs at once, in the open.
Here is why that is the whole test, and why it is harder than it sounds. As long as the two jobs stay fused — as long as the machine can't have it and I've got it arrive together — you never have to check whether one account can actually deliver both. You feel the two together and move on. But they are separable. They answer two different questions. What is wisdom, such that the machine can't have it? — that is a question about the nature of the thing. And do I have it? — that is a question about you. The only way to find out whether your theory can honestly answer both is to make it answer them one at a time, in that order. So that is what we do. We take your answer and ask it the first question, and we let it answer well — because it does. And then, while it is still standing there having done well, we ask it the second. And we watch.
I am not going to do this with some account you do not hold. That would be too easy: I could knock down a theory you never believed and you would learn nothing. I am going to do it with yours. So keep yours in front of you the whole time. When I run the test on my example, run it in parallel on your own answer. I will work the example; the answer is yours.
One more thing before we start, and it matters. Nothing I am about to do takes your answer away from you. The store may be real. The judgment may be exactly and only yours. I am not here to prove you have nothing. Watch what actually happens, because it is stranger and quieter than that: your theory will rule the machine out completely, exactly as you hoped — and by the same logic, in the same move, it will refuse to tell you that you are one of the people who has it. Not no. Just — it cannot say. That is the thing to catch tonight. Not that you lose, but that the account you trusted to do two jobs can only ever do one of them at a time, and the half it does for you is the half you did not need.
Right. Pick it up. Let us ask it the first question.
The first question: what is wisdom, such that the machine can't have it? Ask your answer that, and watch how well it does — because it does well. This is the half that is true, and I am not going to rush it, because you earned this half and it is real.
Say you committed to the act — the judgment, the call you are the one who has to make. Ask why the machine can't have that. What you find underneath, when you say it carefully, is the oldest account we set out, back in the second episode: Aristotle's. Wisdom as the perception of the particular — not knowing more facts, not running a better calculation, but seeing what this exact situation, with these people in it, asks of you, in a way no rule ever reaches, because the wisdom is the seeing of the case, and the case is always particular and always new. That is what you mean, isn't it, when you say the machine does not have to stand inside the choice the way you do. The machine works on the universal — everything that has ever been true of situations like this — and what you do happens at the level of this one, which no amount of the general ever reaches. On that account — and this is exactly what you wanted — the machine is not a rival. It is not even in the running. It can hold every fact about every situation resembling yours and still never perceive the one in front of you, because perceiving is not a quantity of facts; it is a different thing. The information you gathered at two in the morning was never going to settle the decision, just as you felt — not because you had not gathered enough, but because gathering was the wrong kind of act. The machine knowing more than you is not the threat. You were right about that.
Or say you committed to the store, and found that you did not quite mean a heap of facts — you meant what the years had made of you, what living laid down that no reading could. Then you are holding something close to the sage traditions, the ones we met last week alongside Socrates — the Stoics, the Confucians, the Buddhists: wisdom not as something you know but as something you have become, a remade relation to your whole life, slow, earned, lived into. On that account the machine is further out still. It has no life to be remade by. It was never a candidate — not because it lacks data but because it has no self for the data to change. Again, exactly what you hoped. The machine's whole store of information is beside the point, because wisdom, on this picture, was never something you could read your way into, by machine or otherwise.
Or perhaps, when you tested your answer, what came up was closer to Socrates on his own: the sense that the machine's certainty is itself what disqualifies it. A thing that is never once unsure, that answers every question quickly and cleanly and is tired by nothing, is on the face of it the least wise thing in the room, because wisdom may be exactly the knowing that you do not know — the staying open where the machine closes. On that account too you answer the first question without trouble. The machine is completely confident and completely informed, which is, on this picture, exactly the wrong posture. The not-knowing you felt at the table — the very thing that made you feel stupid and stuck — may be the opening the machine cannot find.
Now hold all that for a moment, because I have just handed you something large. Whichever of the three you reached for — the perception, the becoming, the not-knowing — your theory has just done the job you hired it for, and done it completely. The machine, which knows more than any human who ever lived, which out-calculates you on everything you can specify, which sat beside you at two in the morning with everything it knows — on your own account of what wisdom is, that machine is not wise, cannot be wise, is not even the relevant kind of thing. The fear that brought you into this series — that the machine has quietly walked off with the last thing that was yours — is, on your own theory, misplaced. The thing you were afraid of losing was never the machine's to take.
That is real. Keep it. You do not have to hand it back at the end of tonight; the first claim holds, and it is yours. On the account you are holding, the machine is ruled out, and you were right that it is.
It is the second question where the trouble begins.
So ask your answer the second question — the one you have never made it answer out loud. Do I have it? Keep your own account in front of you, because here is the strange part, and I would rather you find it than be told it: the very feature that made your account so good at ruling the machine out is the feature that will not let it confirm you.
Go back to the act — wisdom as the perception of the particular. That account ruled the machine out by saying wisdom is not a quantity of anything, not facts, not calculation, but a seeing that no rule reaches. Good. Now: what is the test? How do you check whether you perceive the particular, or only think you do? And the account has no answer — not by oversight, but by its own logic. There can be no exam, because an exam is a set of rules, and the whole claim was that wisdom is what no rule reaches. Aristotle is clear about this: the clever young person can know every principle and still lack it, and you cannot tell the one who has it from the one who merely argues well by any test you could write down. The only thing that ever shows it is the long, slow fact of whether you actually acted well — across a whole life, in the particulars, again and again — and that is exactly what you cannot read off from where you sit tonight. Other people, later, with the whole record in front of them, may see it in you. You, here, wanting to be wise, get no view of it at all. The same thing that rules the machine out rules you out. It was never a boundary around you. It was a boundary around wisdom itself, and you happen to stand outside it too.
Now run it on the becoming — the sage. That account ruled the machine out cold: no self, no transformation, not a candidate. So check yourself against it. Have I done the transforming? Am I the remade one? And every one of those traditions says the same hard thing back: the question disqualifies itself. The sage does not wonder whether they are a sage. The one who is sure they have arrived has shown, in being sure, that they have not — because part of what you become, if you become it, is someone who has stopped needing the confirmation. The traditions agree, and they agree on purpose: you get no privileged view of your own attainment. So the account that put the machine furthest out of the running puts you out of the running for the confirmation in the very same sentence.
And Socrates is the sharpest of the three, because with him the bind is tightest. That account ruled the machine out by saying wisdom is knowing that you do not know, and the machine's confident omniscience is its opposite. Good. Now you go to claim the confirmation — so then I, who feel my not-knowing, I must be the wise one. And the account denies it to you. Because the moment you are sure you have it, you have taken the exact posture — confident, settled, certain of the answer — that the account calls the surest mark of not having it. You cannot claim this one without losing it in the claiming. The machine can't have it because it is too sure. And you cannot confirm you have it, because confirming would make you too sure. The thing that disqualified the machine disqualifies your claim by the very same rule.
So there it is, and notice that you did this, not me: every account strong enough to rule the machine out is, by the identical logic, too strong to confirm you. The boundary you drew with yourself safely inside it turns out never to have been drawn around you. It was drawn around wisdom, and it keeps out everyone who tries to point at themselves and say: mine, confirmed. The machine and you, on the same side of it.
Now — I can hear the other objection coming, because I raised it myself the first time I ran this. You feel the accounts that exempt human beings refuse you the confirmation, and something in you says: fine, then I will take an account that does confirm. Do not tell me there is no test. There is a test, and I can point to it — the hard calls I got right, the times I saw it coming when no one else did, the track record. I can show my work. And you can. There are real accounts that say exactly that, and we met both of them in the third episode — the two that treat wisdom as a competence you build and can check. One says your good judgment is a high skill — the expert's feel for a whole situation, the thing the seasoned nurse and the chess master have, earned across a thousand lived cases — and, like any skill, it shows in performance: the proof is the record. The other says it is a deep, checkable rationality — your commitments hanging together and tracking what actually matters — the kind you really can audit from the inside, fallibly but honestly, by your reasons. Either way, here at last is an account that hands you the very confirmation the others refused. Reach for it. It is there. It is yours for the taking.
But notice what you actually get when you take it. This confirmation is checkable — that is the whole point of it, the only reason it can finally say yes, you, confirmed. And a thing that can be checked can be checked on anyone, and on anything. If wisdom is a skill built from vast experience, then a system trained on enough of the same cases is a candidate to have built it too — and that was the whole live argument in the third episode, where even the man who first set out what expert skill is bet that it would keep machines out, and that bet is exactly what is now in doubt. And if wisdom is reasons that hang together and a record that tracks what matters, then a system that can lay out its reasons, show its calls, and be audited the way you would audit yourself is just as much a candidate. The exact feature that lets these accounts confirm you — that wisdom works by something you can point to and inspect — is the feature that makes the machine a candidate again, at the very point where you had ruled it out. You wanted the confirmation and you wanted wisdom to be yours alone, and these accounts give you the confirmation only on one condition: wisdom is no longer yours alone. You let the machine back into the running, and you let it in because you reached for the proof.
So look at where you are now, with your own answer in your hand. Reach for the accounts that rule the machine out, and they will not confirm you. Reach for the accounts that confirm you, and they will not rule the machine out. The two halves of what you came in holding — the machine can't have it, and I've got it — will not both stay in your hand at once. Every account that gives you the first takes the second; every account that gives you the second gives it to the machine as well. And nobody did that to you. The accounts are all still standing, every one of them, exactly as serious as they were an hour ago — none defeated, none the winner. You simply asked your own theory to do its two jobs at the same time, in the open, and watched it refuse.
Here is what you take out of tonight, and it is not an answer — it is a sentence you can no longer say in one breath.
You walked in holding one fact: the machine isn't wise, and I am. You had carried it your whole life as a single thing — smooth, fused, the obvious truth of someone sitting up at two in the morning with a hard call and a clever machine that cannot make it. And tonight you examined it closely and found it was really two claims, not one. The first piece is true, and on the accounts that rule the machine out of wisdom altogether — the seeing of the particular, the becoming, the not-knowing — it is yours to keep: there, the machine that knows more than you is no rival, and the thing you feared losing was never its to take. You can stop being afraid of that. The second piece — and I am — you reached for it tonight through every account you hold, and not one of them would put it in your hand. The accounts that rule the machine out will not confirm you. The accounts that would confirm you will not rule the machine out. The thing you were proudest of — your judgment, the call only you can make — turns out to be exactly where the confirmation is not. Not because you lack wisdom. Because the kind of thing wisdom is, on your own account of it, is the kind of thing no one gets to confirm they have from the inside.
And I am not going to soften that into something comfortable, because the whole point of this series is that I will not sell you the easy version. There is a consoling way to hear what just happened — well, at least now I see it clearly, at least I am too honest to be fooled — and I would ask you to notice that this is only the second claim again, coming back in a new form. At least I see it clearly is one more way of saying and I've got it, with clarity standing in for judgment this time. You do not get out of this by being the one clever enough to describe it. It is built into the thing itself. You leave tonight able to rule the machine out of your judgment and unable, by the same logic, to confirm there is any judgment of yours in there to keep safe. That is not a loss of nerve on my part, and it is not a trick. It is what your own theory says, the moment you make it state both halves at once.
Now, there is one more place this goes, and it is the one that will cost the most, so tonight I am only going to point at it. You have spent six episodes learning to do a particular thing: catch the buried theory, hold an answer that will not resolve, refuse the easy version, and stay with a question instead of forcing it shut. And I would guess you have begun, somewhere, to think of that capacity as a kind of wisdom — perhaps the kind you have been building here, week by week. So next week we ask the question I have been holding back: that capacity — the one this whole show has been training into you — is it wisdom? Or is it only something that looks like wisdom? I am not going to tell you tonight. Next week we turn the test on the show itself.
But carry one real thing into the week — an actual moment, not an invented one. You will do the reflex again: someone will say the machines have this now, they will be wiser than all of us soon, and you will feel the no come straight up, the same as ever. Let it. You are allowed your no; tonight earned you the first half of it honestly. But this time, finish the sentence all the way to the end that the account licenses. Not there is a thing the machine can't have, and I've got it. Instead: there is a thing the machine can't have — and I have no way, from in here, to tell you I have got it either. Say the whole sentence. Both halves. Notice how different it is to carry the true half without the flattering half attached to it. That is the whole exercise: not to lose the no, but to stop it smuggling the yes.
Thanks for listening. I'll see you next time.