 All right, good morning, everyone. I know we're a little bit behind schedule. I'll try to get us back on track. Because essentially, I don't have much to say here. The title of this talk that I want to give right now is, Do Big Data and AI, Solve the Economic Calculation Problem. Well, I have Dr. Lamber. He was my co-author on this paper. Of course, it's me here presenting it, but I feel like he needs a shout out because he did most of the work. But you have me presenting it. If you have any questions, ask him. So the impetus for this paper was a special issue of the Journal of Economic Behavior and Organization. They had this special issue with the title Machine Versus Market, Economic Calculation and Business Organization in the Age of Modern Computing. And the question of the special issue was, will modern computing techniques pose a fundamental transformation of economic organization? And basically, what we had to say was, if you read and understood Mises, you wouldn't even be asking this question. So basically, the point of the paper was like, here's what Mises said. Modern technology hasn't changed that. So no. So that's why I say this will be short. And we can get to lunch a little early, maybe. But I will summarize the argument. It will retread a lot of ground that Dr. Salerno talked about in his lecture on economic calculation and socialism, and then in light of Big Dad and AI. So here's our paper. It does exist. It was published within the last year. So feel free to read it. OK. So this problem of economic calculation. So just to find, what is economic calculation? The calculation part is simply arithmetic. You have numbers, you add and subtract them. That's calculation. The economic part or the economizing part is you have scarce means. You want to allocate them to your most highly valued uses. You don't want to allocate them to less valued uses. You don't want to forsake those higher valued uses. That is economizing. So like our first day lectures talking about subjective value, we learn that value is ordinal rankings. You prefer this to that. You can't do arithmetic operations on this. You can't say with my first sack of grain that I like that five times as much as that fifth sack or anything like that. It's just more to less. It's ordinal. It's not cardinal. You can't do arithmetic operations on it. We also have producer goods. So as Dr. Slavo mentioned, you can make production functions. And say this producer good requires, so if we have this producer good H, we'll call it H, we have various combinations of productive factors we could use to produce H. Maybe four units of A and two units of B or three units of A and six units of C or five units and three units of F. But I mean, this is just a technological relationship. It's not an economic one. We want to know the opportunity costs of A and B and C and F in order to determine what combination of factors to use to satisfy our most highly ranked ends and what the opportunity costs of those factors. What are we giving up by choosing one of these combinations compared to others? So the problem is, I mean, before we get to the calculation part is we need some way to arithmetically compare means and ends to mentally ascertain whether we're using resources in an economizing way. If all we have are like these, if we know all these physical resources we can use to produce things and we produce these things. We don't know what we're giving. We don't have this common unit in which to compare our yield with our means. So again, starting with just ordinal values, we can't, just based on those alone, we can't calculate. We need a cardinal unit. There's no cardinal unit naturally available to determine the effectiveness of action. By effectiveness, I mean, are we allocating resources to more highly valued uses? So how do we move from ordinal subjective valuations to precise calculation of profit and loss? We need the market economy or the social appraisement process, as Dr. Salerno mentioned. So the social appraisement process, producers value producer goods according to anticipated consumer valuation. So if you recall from Dr. Klein's lecture, entrepreneurs anticipate future consumer prices and their willingness to bid on various factors of production is based on what they expect, the discounted marginal revenue product of those factors to be in certain lines of production. And they need prices in order to understand that current consumer prices, they're a help in anticipating future prices, but you can't just take current prices and know what future prices will be. There has to be that entrepreneurial appraisement of the future. And so since consumer goods are exchanged for money, consumers demonstrate they value a consumer good more than the money they give up for it. And entrepreneurs are creating this price structure for the higher order goods when they're bidding for factors of production in anticipation of the future prices of goods that they will create at the end of the production process. This creates this monetary structure that has a commensurate money unit or monetary price for both consumer goods and producer goods, making calculation possible. So entrepreneurs, there's both this backward-looking process and forward-looking process where after they've engaged in a productive process, they can ascertain whether they've made profit or loss. They paid certain prices for producer goods and they created consumer goods. They sold those. They can engage in this cost accounting. And as they're, and there's always also this forward-looking aspect, so backwards forward. I'm sure. Say this is backwards, this is forwards. They are able to, they anticipate future prices. Bid on, as I mentioned, the current factor. So factor prices have baked into them these anticipation of future prices. So that's the forward-looking aspect. Of course, entrepreneurs might make errors. They make losses. And some socialists will criticize, look at this economic calculation. You think it's so important. It's not perfect. It's kind of like, this is the point. We can tell that there's loss, right? That's not a failure of the system. That's the system working, right? Because under socialism, you don't know whether you're making profit or loss. You don't know whether you're allocating resources to more highly valued uses. And I think Dr. Salerno says it well. He's worth quoting, not too lengthy here, but this is from the epilogue of the Mises Institute edition of the 1920 article, Economic Calculation to Socialist Common Wealth. He says, there thus comes into being the market's monetary price structure. A genuinely social phenomenon, and I like that social part. It's really ironic about socialism, is that in the market economy, every market participant has an input, can affect this, is involved in this genuinely social phenomenon. Another social is just the central planner. So it's not very social to that extent. So they create this genuinely, or at least this genuinely social phenomenon, which every unit of exchangeable goods and services is assigned a socially significant cardinal number. So that's also important, that it has social significance. This is some of the cyber communists. I mean, it's not pejorative terms. That's like what they're actually called. In their proposals of how do we overcome this problem that Mises pointed out? I mean, they try to come up with some kind of cardinal numbers, but they're not socially meaningful. They're often just arbitrary, and I'll talk about some of those. So you get this socially significant cardinal number, which has its roots in the minds of every single member of society that must forever transcend the contribution of the individual human mind. So that's this social praisement process where you create this unified price structure, incorporates consumer valuations of consumer goods and entrepreneurs' anticipations of future prices and affecting how they bid on producer goods. So you have this meaningful way to compare inputs and outputs in monetary terms within a meaningful cardinal unit. And since I'm quoting, I have a big Mises quote. I won't apologize though. I think Mises is worth, always worth quoting at length. What did I notice in, so it was really nice with this special issue. I had the opportunity, the editors of it, organized kind of a paper workshop where contributors could get early views on drafts of people's papers and trash them. And what I found, like communicating with the cyber commies, is that there's this disk, they don't really incorporate value theory. And as Dr. Salero mentioned, this has been a problem for a long time. Even among the classical school of commerce, they weren't able to effectively respond to the socialists because of their flawed value theory. So what Mises says here is that, all attempts at disproving the conclusiveness of my thesis, and I love reading Mises when he's talking about his 1920 paper. Sometimes he does it in the third person. It's like this irrefutable smash of the intellectual case for socialism. So I mean, the conclusiveness of my thesis were destined for failure because they did not penetrate the value theory at the core of the problem. All these books, theses and essays tried to rescue socialism. They wanted to show that it was indeed possible to construct a socialist commonwealth in which economic calculations could be performed. They failed to see that one must begin with the question of how in an economy consisting of preferring and deferring that is making unequal valuations, one can arrive at comparable valuations in the use of equations. And this goes back to Dr. Jonathan Newman's lecture, how neoclassicals try to do this with indifference. They're trying to create a quality, but you can't really do that when you have. Remember, we start with ordinal preferences. It's inequalities, but they try to smash it into equations rather than inequalities. So it was that they came upon the absurd idea of recommending the equations of mathematical catalactics which depict an image devoid of human action as a substitute for the monetary calculation of the market economy. And they continue to do so. So in addition to this problem of not, you gotta really start with value theory in understanding this calculation problem. Another issue and a related issue with the socialists, the cyber communists is this confusion of an economic problem versus a technological problem. So if I, you might see this even in a lot of economics textbooks where you have an objective function and say you're trying to maximize something subject to these constraints. Well, this isn't an economic problem, this is an engineering problem. So if the socialists, the planner, specifies certain ends and wants to achieve this end, like maybe minimize the resources used to achieve this end. Now this is not dealing with economizing because it's not considering the alternative uses of those resources. It's just a technological or engineering problem. So to quote Mises again, says technology operates with countable and measurable quantities of external things and effects. It knows causal relations between them, but it is foreign to their relevance to human wants and desires. It feels as that of objective use value only. That is creating physical goods. We can know that, that's a technological thing, that knowing technical rates of substitution, what physical goods you can create with what physical inputs. But it's not relevant to human wants and desires to value. So gotta say a simple example to illustrate this difference. So suppose a central planner is magically endowed with knowledge of all quantitative causal relations. So knows all the technology involving resources under his control and only cares about his own well-being. So even setting aside the issue of producing for everyone, he just wants to satisfy his own ends. So let's say one of his ends. He wants to build a bridge to nowhere as government actors like to do. So he wants to build this bridge and he has to choose among various resources available to him to decide what resources to use to build this bridge. So the question is, how does he select that combination of producer goods that yields the bridge most economically with respect to alternative uses of those inputs? So there's a very simplified example where you can choose between a platinum bridge or a stone bridge. And all it takes is to build this platinum bridge, A number of units in kilograms of platinum, some B units of labor and some C units of land area. Now this thing about this, I was reminded of that Monty Python and the Holy Grail with the anarcho-syndicalists where they're just playing in the mud, we just need land and labor and that's our production process. The production process for the stone bridge is very similar. As you can see, it also requires B units of labor and C units of land to make this very simple. The only difference in inputs is D units of stone. So here, how does the planner choose between these? How can we compare A plus B plus C, each of which are in different units? We've got a weight, links of time, land area, with P, P being the bridge, the finished bridge, or the other production process where it uses D units of stone and the same amount of labor and land with S. Or can we compare A with D? Can we compare kilograms of platinum with kilograms of stone? I mean maybe if we want to have a lighter bridge, I don't know. But that's not economically relevant. So if we try to do something like that, so claiming P minus A, I mean it's nonsense, we have one bridge and we subtract A kilograms of platinum, we say, oh that's less costly or less whatever than one unit of a stone bridge minus D units of stone. And therefore the stone bridge could be construct instead of the platinum bridge, it's not meaningful, we're using incommensurate units. So there's really no way that this planner, even producing for his own consumption of this bridge to nowhere, there's no calculation that's possible, where unlike if we did have a market for platinum bridges and for labor and land, there could be carnal numbers. Instead of a P, P would represent a monetary price and so would A and S. So we'd have this commensurate unit in money prices in which we could meaningfully calculate profit and loss and whether it's more economically effective to, but I mean, are we allocating resources from our highly valued uses compared to their opportunity costs between the stone bridge and the platinum bridge? And what I find notable, so kind of a similar example in human action, Mises' comment when he goes through a thought experiment like this is again quoting Mises, says today we calculate from the point of view of our present knowledge and of our present anticipation of future conditions. We do not deal with the problem of whether or not the director or planner will be able to anticipate future conditions. What we have in mind is that the director cannot calculate from the point of view of his own present value judgments and his own present anticipation of future conditions, whatever they may be. So it's really, I gotta say damning here to the socialist case, because he's like let's set aside conditions or anticipations of future conditions. Just looking at what stone and platinum we have now, I mean, the planner's not even thinking about maintaining his capital stock, not that the capital stock would be meaningful, but capital goods, it's a really intractable problem, even not even considering the greater complexities of having to plan for the future. Even in this very simplistic case, the planner is unable to engage in calculation. So I was thinking of the market economy, there's this difference in kind between a monetary economy and a non-monetary economy. Value calculation, it's impossible in a society without money. You need that commensurate unit. It's really fundamental to calculation. And if you recall from Dr. Rittner's lecture on Austrian capital theory, this category of capital only exists in a market economy. You can have in a socialist economy, or not in, I guess I shouldn't call it a socialist society, there's no economizing, right? Under a system of socialism. You can have capital goods, but you don't have capital because capital is that sum of the monetary values of capital goods. You can only have that when there's exchange of capital goods. So this category of capital only exists within a market economy. And as Dr. Salano emphasized that there's no economy because capital accounting is impossible under socialism. There's no economy in terms of rationally allocating resources according to their most highly valued uses. Okay, so just really reiterate this problem facing the central planner. I mean, it's fully possible for the central planner to specify an end and then organize the factors of production in order to attain that output. But it's unable to economize on the employment of those higher order goods, according to their genuine opportunity costs. So in the words of Mises, the central planner will be floundering in the ocean of possible and conceivable economic combinations without the compass of economic calculation. He had a lot of metaphors for how, or another one, the central planner would be groping in the dark. And I think this is actually a bit too generous of Mises, because you think like, if you're groping around in the dark, you might actually find what you're looking for. But the thing about the socialist planner is they can't know. He can't find what he's looking for. So do all the groping he wants, but it won't find the, sorry, the most economically efficient means. So it really seems like Mises is right, he really smashed the case for socialism. And it's still today. It says, well, of course, some stubborn Marxians raised objections, and they still do. So it's a difficulty in talking about this, okay, so we now get to this big data and AI change things. And it's like, well, it's hard conceiving, like how could they? So that was a bit of a puzzle, like okay, what would it mean for them to solve the calculation problem? Well, I guess for, it's still to be socialism, you'd have to get rid of private property and the means of production. So specifying the institutional prerequisites for economic calculation, one of them being private property and the means of production, free exchange of those goods. And secondly, a common medium of exchange, money. So you have free exchange, money prices, I mean, capitalism. Here you have these things, and these are the things you need for economic calculation. So any argument that technological advances can overcome the calculation problem needs to provide a substitute for the functions that private property and money serve, okay? So it must overcome certain hurdles. Anything else you're doing, it's a side issue, which some of the socialists do, is that you're not addressing the calculation problem, it's not occurring. So these hurdles to overcome are these things we get from private property and money. So first, commensurate cardinal units that are meaningful for economic decision-making, considering the economically efficient employment of scarce resources. And then some mechanism from transforming ordinal rankings to commensurate cardinal numbers. Because if all you have is ordinal preferences, you can't do arithmetic operations. You have to have some way of turning them into commensurate cardinal numbers. And thirdly, some capability of forecasting future arrangements of market data, because this, as we'll see, is a big problem with big data. It's all, how do you use it to anticipate future market arrangements? I mean, not only of amounts of current consumer goods, but novel new consumer goods. And how do you know whether those investments are worth pursuing? So, this question, can big data overcome these hurdles? No, no, that's an inclusion to come to. So for one thing, the data are entirely that of the past. It's not clear how you can just take that and magically turn it into a future price. You still need that entrepreneurial function of appraisement, of anticipating future prices. There's nothing forward-looking about data from the past. And while we acknowledge that, I mean, these technologies might be useful for entrepreneurs operating in a market economy. Say, maybe Amazon, they got all their data, they have these huge inventories, maybe big data, whatever big data is, helps them keep up-to-the-minute tabs on what they have with their inventory records and technology might help us better understand technical substitutions, but it doesn't take these and turn them into market prices. So, it really doesn't address this problem. Dr. Herbner says, said a number of years ago, this is still true, you got apples and oranges, no advances in supercomputers can overcome the impossibility of adding them together. They're just incommensurate. So, what about AI? The analysis is similar in most ways. Again, just as big data pass data can be valuable for entrepreneurs, it's imaginable that the use of AI can help entrepreneurs in various ways, but it's not clear, just as Misa said, let the central planner have all knowledge of ordinal preference rankings, all the resources available to him, all the technical rates of substitution and production possibilities. Still can't create a price structure, so it's not clear how AI can do that. There's further issues, so an editor and contributor to this volume of the Journal of Economic Behavior and Organization, C. Phalen, he brings up this issue of whether AI can create this theory of mind in order to predict human behavior. Maybe Dr. Gordon can predict, successfully predict, but it's not clear how AI can do it. It seems like there needs to be some, yeah, some frame of mind, or a theory of frame of mind. It's also unclear whether AI can break frame and innovate out of specified parameters, because a lot of the cyber-communist models, I mean, they start with pre-specified outputs, but there's no room for them in creating novel goods, which, I mean, isn't important issues. I mean, if you have a stationary economy, maybe it'd be less of a problem, but that's not what socialism promised us, right? We promised me more. So yeah, again, doesn't seem like, oh, I tried this. So I went, maybe I can't read this. So I told the chat, asked it to centrally plan an economy, and I don't know what I expected. Maybe like, it lists all goods and the quantity and prices they should be produced for each day. It'd be a lot of output, but this is what it told me. It's like, oh, designing a centrally-planned economy, it's evolved central authorities doing these things. It's been attempted, but it has drawbacks and challenges. So here's all the things you have to do. It's a lot of things. I mean, the list keeps going on, but like number two, it says resource allocation. The central planning authority must determine the allocation of resources among various sectors and industries. This could involve assessing the country's needs and priorities, could. And distributing resources accordingly. So I go, okay, how do I do that? So maybe you can try a different command and it can give you a more useful output, but I don't know how to use AI right. I addressed this question. I found it interesting that Dr. Salerno and mentioning some socialist responses to MISA, some of the early naive ones included using labor time as this commensurate unit or cardinal unit, incommensurate cardinal unit, and they're still doing it. So one of the papers we criticized, one of the more recent, well, additions to this literature, uses labor time as a basis for economic education. Again, the answer is no, they can't do it. And why not? Well, there's many flaws. The most fundamental one is it doesn't allow you to economize on the produced factors of production. You just can economize on labor. And as far as I can tell from their model day, just like the Marxists of old, they see capital goods as embodying this labor. So I think they assign the hours of labor that it took to produce this capital good, but it's not clear how this is useful in terms of considering the opportunity costs of these factors of production. It took this many hours to create it, and I don't think they adjusted for different production processes. So it's just, they got a number, but it's not meaningful in terms of determining genuine opportunity costs. So one example of a paper that, or actually this was a book by Cockshot and Cattrell, Cockshot was a contributor to this Gibo issue. They start with these planned targets. They use mathematical optimization to meet the targets, given constraints. And then, so like, they acknowledge like, okay, yeah, we wanna incorporate consumer demands and adjust production based on them. So I actually really need to reread it, because as I understand it now, I'm like, this doesn't, how does this work? So because they issue tokens to laborers, like, I don't think it's like, you get this labor hour, do you work an hour? Here's your one token. You work two hours, here's your labor token. And this, how they initially set prices. I'm like, do they do it based on how many hours it took to create this good? So I'm like, how can anybody afford anything? Like say the two of you, say you and your coworker, you've both spent 10 hours, so 20 hours total producing this good. So it costs 20 hours. But you'll both get 10 tokens, like neither you can afford it. So I don't know how, maybe the pricing process is different. So go check it out yourself. If you're interested in such a thing. But after this process of offering these consumer goods and pricing them in some way, and observing consumers spend their tokens, then they're gonna adjust future production based on this, any divergence between the assigned price and how many consumer demand in terms of these tokens. But however they do this initial pricing process, this procedure has nothing to do with economic calculation. They do this initial allocation of resources. It pre-specifies what ought to be produced. It doesn't have any necessary relation to what consumers want. But again, this is like described before, this is a technological problem they're doing, not an economic problem. And this feedback mechanism where they're saying, oh, we have consumer goods markets, we're gonna adjust production based on it. Yeah, and they don't consider producing the production of alternative goods, it's just, here's our list of goods that we initially produced. And I guess that's what we're gonna be consuming through the rest of time. Some other cyber communists. So they acknowledge it. So Philip DePreech, another contributor to this volume. They seem to acknowledge this problem with the labor theory of value. Yeah, he actually says, oh, yeah, Bumbavirk was right. Good, we're making progress. But these trial and error methods, go read chapter 26 of Human Action. I mean, Mises talks about, to have trial and error, you have to realize you're making errors. And you can't do that without economic calculation. So DePreech acknowledges this, but then he's like, okay, we're not gonna use labor, we're gonna use shadow prices. But the resulting figures say, they aren't capturing genuine opportunity costs, they're just arbitrary. Another interesting observation made by Nieto and Matteo, like they're like, look at Walmart and Amazon. They're big, why don't we just have them in charge of the, oh, I don't, they don't notice they're like, let's have Amazon just run the economy. But yeah, they're big, but they make use of market prices. They couldn't do what they do without the use of economic calculation. So it's not really a legit response. So my, see this is like ending with reading recommendations are popular. So read Mises. You get something more out of each time going through Human Act, just realizing the depth of his contribution regarding economic calculation, in addition to other things, of course. Don't confuse economic problems with technological ones. And I'll end with just a couple more Mises quotes, as Dr. Thorne calls them, the quotable Mises. The measurements of physics and chemistry make sense for practical action, only because there is economic calculation. Is monetary calculation that made arithmetic a tool in the struggle for a better life? Provides a mode of using the achievements of laboratory experiments for the most efficacious removal of uneasiness. So it's like, yeah, it's great you have these technological improvements, but it's really economic calculation that allows you to make the best use of them. The closing quote, so I'm finally done. I'll stop quoting Mises, at least for now. Which I like this quote. Says, our civilization is inseparably linked with our methods of economic calculation. It would perish if we were to abandon this most precious intellectual tool of acting. So these are not reading recommendations, but I figure I should list my references of the work cited. Thank you for your attention.