Gift vs. reputation in hacker culture

A G+ follower pointed me at Note on Homesteading the Noosphere by Martin Sústrik. He concludes saying this:

In short: Labeling open source communities as gift cultures is not helpful. It just muddles the understanding of what’s actually going on. However, given that they are not exchange economies either, they probably deserve a name of their own, say, “reputation culture”.

I’m going to start by saying that I wish I’d seen a lot more criticism this intelligent. It bothers me that in 20 years nobody seems to have refuted or seriously improved on my theories – I see this as a problem, both for the study of hacker culture and in the field of anthropology.

That said, I think Sústrik gets a couple things wrong here. And don’t want them to obscure the large thing he’s gotten right.

First (possible) mistake: I have not observed that, as a matter of language, the term “gift culture” is as hard-edged and specific as he thinks it is. There’s a way we could both be right, though – it might be that terminology has shifted since I wrote HtN. Possibly this came about as part as the revival of interest in the concept that I seem to have stimulated.

But: one piece of evidence that anthropologists are still using “gift culture” in the inclusive sense Sústrik criticizes me for enmploying is that Sústrik himself feels, at the end of the article, that he needs to propose a contrasting term rather than citing one that is already established.

This so far is all about map rather than territory. As a General Semanticist I know better than to get over-invested in it.

Here’s the territory issue: Sústrik is not quite right about expectations of direct reciprocal exchange not being a shaping force. True enough that they aren’t salient at the macrolevel the way they were among the Kwakaka’wakwe. But if I download a piece of open source, and it’s useful to me, and I find a bug in it, I do indeed feel a reciprocal obligation to the project owner (not just an attenuated feeling about the culture in general) to gin up a fix patch if it is at all within my capability to do so – an obligation that rises in proportion to the value of his/her gift.

I should also point out that the cultures Sústrik think are paradigmatic for his strict sense of “gift culture” are mixed in the other direction. There is certainly an element of generalized reputation-seeking in the way individual Kwakaka’wakwe discharged their debts. There, and in the New Guinea Highlands, the “big man” is seen to have high status by virtue of his generosity – he overpays, on the material level, to buy reputation.

In the terms Sústrik wants to use, open-source culture is reputation-driven at both macro-level and microlevel, and also sometimes driven by gift reciprocation in his strict sense at microlevel. The macro-level reputation-seeking and micro-level gift reciprocity feed and reinforce each other.

This brings me to the large thing that Sústrik gets right. I think his distinction between “gift” and “reputation” cultures is fruitful – both testable and predictively useful. While I’m still skeptical about it being in general use among anthropologists, I rather hope I’m mistaken about that – better if it were.

Yes, real-world cultures are probably never pure examples of one or the other. But differentiating the mechanisms – and observing that the Kwakaka’wakwe and hacker culture are near opposing ends of the spectrum in how they combine – that is certainly worthwhile.

As a minor point, Sústrik is also quite right about reciprocal licenses being a red herring in this discussion. But I think he has the reason for their irrelevance mostly wrong. The important fact is they’re not mainly intended to regulate in-group behavior; they’re mainly a lever on the behavior of outsiders coming into contact with the hacker culture.

(It was actually my wife Cathy – a pretty sharp-eyed observer herself, and not coincidentally a lawyer – who brought this to my attention.)

Bottom line, however, is that this was high-quality criticism that got its most central point right. In fact, if I were writing HtN today I would use – and argue for – Sústrik’s distinction myself.

58 comments

  1. Thanks for posting this.

    A little while ago, from what I knew of hacker culture, and from what I had read about Moka exchange, I had this nagging feeling of, “They’re not quite the same thing,” but I couldn’t pinpoint the distinction in my thoughts.

    This seems to happen often enough to deserve a name. Is there a standard term for the feeling that X is a decent model of Y, but somehow fails to capture a subtle (and in principle, testable) difference?

  2. How does that apply going the other way – from maintainer to semi-drive-by contributor?

    The case I experienced was receiving a nicely-encapsulated, beautifully-explained patch that also inadvertently filed a couple of project-level bug reports (like me not having made contribution policies explicit – such as not dropping coverage, testing new changes, don’t stop the music, etc). I jumped on the patch like a hobo on a ham sandwich and heartily encouraged the bloke who sent it to keep ’em coming. (He since has, with a bug report of similar quality, and an insightful question to follow up with).

    The initial patch had a bug in it (I can’t point fingers – I’ve made that exact goof multiple times), which I pointed out, and suggested a way to both fix bug, and verify that fix. I was quite reluctant to make those changes myself, but ultimately had to bolt the fix on top of his patch and roll the combination up – I didn’t want to leave a good patch languishing, and made sure the original contributor got their name in lights in the changelog.

    Yes, part of this was self-interest (not wanting to chase future contributors away pre-emptively, or end up looking like a goose). Or am I well into “all interesting behaviour is overdetermined” territory here?

    1. >Or am I well into “all interesting behaviour is overdetermined” territory here?

      Yes. :-)

      A thing that happens with social/altruistic behaviors like this is that people internalize them and experience doing them as personal satisfaction. I think this is a way to increase your perceived trustworthiness; humans are good at cheater detection, so the safest strategy is to actually be reliuable (or, at least, defect seldom enough not to get caught doing it).

  3. I thought so, ESR – thanks for confirming that.

    Is anyone else misreading the title as “Git vs reputation….” ? Had the record scratch a couple of times before I twigged.

  4. Status/reputation is basically a currency, http://www.meltingasphalt.com/the-economics-of-social-status/

    “We trade status for favors (and vice versa). This is so common you might not even realize it, but even the simple act of saying “please” and “thank you” accords a nominal amount of status to the person doing the favor. The fact that status is at stake in these transactions becomes clear when the pleasantries are withheld, which we often interpret as an insult (i.e., a threat to our status).”

    I don’t know much about hacker culture but these status-transactions don’t really seem to be happening. The basic idea of a status transaction would be that some young, inexperienced fellow spends months doing some work like testing for free and then Linus publicly thanks him, he gains a lot of status by that, and Linus loses a little bit of status although that is a little bit obscure how and why (but imagine if he had to do that all the time, 24/7, you can’t just keep throwing status at people without losing some of your own), then the young guy leverages that into getting a job, or consulting gigs, exchanging status for money.

    But is this really happening? But if not – if reputation is not used transactionally – how exactly is it a reputation _economy_, instead of “just” being a reputation _culture_ ? Not everything that is productive is an economy, it is transactions and exchanges, not productivity, that define economies. Highly productive cultures do exist and I suspect that is exactly what you guys are having.

    1. >and Linus loses a little bit of status

      No, this is completely wrong. Linus gains status by been seen to praise good work – by acting as a leader and a big man (in the sense the Moka of New Guinea use the term) should. Big men are expected to reinforce gifting norms and at the same time show generosity towards those of lesser status.

      >But if not – if reputation is not used transactionally – how exactly is it a reputation _economy_, instead of “just” being a reputation _culture_ ?

      Because the your reputation drives how willing other people are to give you help on your projects. It’s a kind of intangible capital that cashes out in other peoples’ labor.

      One of the principal benefits of being “ESR” is that when I spin up a new project people show up almost immediately to help me raise the barn. Often within a day or two of the launch. You get this by having high status; also be being known to be scrupulous and generous about crediting others.

  5. >A thing that happens with social/altruistic behaviors like this is that people internalize them and experience doing them as personal satisfaction. I think this is a way to increase your perceived trustworthiness; humans are good at cheater detection, so the safest strategy is to actually be reliuable (or, at least, defect seldom enough not to get caught doing it).

    Internalization begins right at how we raise children, we lightly shame them if they something bad, praise them if they do something good, hijacking their desire for respect i.e. status in the eyes of parents (and it is respect, not just love, “I love you even though you are the worst kid in the neighborhood” is not exactly what kids like to hear), which turns into an internal “policeman” called conscience, that proceeds to self-shame when we do something bad, and gives a warm glow when we do something good. It is not exactly a new idea, Freud was driving at this with the “superego”.

    It is precisely due to this internalization process why status became a really popular way to explain human behavior, see Overcoming Bias. It is not that we are all a bunch of Machiavellians, it is that we all like to feel good about ourselves, we want to feel we are good, not bad, and the two are closely related. Here is how they are related:

    https://en.wikipedia.org/wiki/Sociometer

    I must admit I did not take the concept of self-esteem entirely seriously until I learned that the CBT way of fixing depression is mostly self-esteem boosting (“learn that not everything always your fault” etc.) so depression can be defined as chronically low self-esteem, so self-esteem is something people can literally kill themselves over.

    If sociometer theory is right and the two can be linked i.e. low self-esteem is self-shaming, internalized low status and high self-esteem is self-praise, internalized high status, we have a nearly completely model: we just want to feel good about ourselves, an entirely understandable, non-Machiavellian way, but our opinion about ourselves is learned from how others form opinions (“Mom and dad think liars are disgusting”) which ends up in practice at least halfway like a competition for status.

    And this is actually how Rene Girard’s whole mimetic desire stuff can be made to make sense. As without this it has no legs really. Girard says others think X is a great artists so we learn to like him, in reality others think everybody who considers X a great artist has good taste, and we want to feel we have good taste, we want that for our self-esteem, so a subconscious process makes us like that artist.

  6. @Alex

    Well ‘git’ is a British word translating roughly to “jerk” or “asshole”, and Linus said that he named it after himself. So I guess git would be a relevant example of reputation.

  7. >>and Linus loses a little bit of status

    > No, this is completely wrong. Linus gains status by been seen to praise good work – by acting as a leader and a big man (in the sense the Moka of New Guinea use the term) should. Big men are expected to reinforce gifting norms and at the same time show generosity towards those of lesser status.

    That it’s completely wrong is basically his point: He’s saying that Linus losing status is what would happen if the interaction described were a “transaction”, and if the interaction is not a “transaction” then the entire pattern of such interactions across hacker culture is not an “economy”. I think basically what he means by “transaction” is “interaction involving the allocation of a rival good”. Obviously, Linus doesn’t lose status in the interaction, both parties gain status, so we can see that status isn’t a rival good, so this interaction isn’t a “transaction” and thus isn’t part of an “economy”.

    I would argue, though, that there is a rival good being exchanged here: the time of the parties involved. The young whippersnapper devotes time to hacking on the kernel, making bug reports, analyzing the performance of a new feature, etc. Linus then devotes time to intergrating the results of that work into the kernel (merging a patch, squashing a bug, making a decision based on performance data, etc). The problem Linus has is that some time investment is required just to determine if the contribution from the whippersnapper is worthwhile, so if the whippersnapper’s contribution is worthless (buggy code, or code that doesn’t contribute anything meaningful, a bug report of the form “FEATURE NOT WORKING!!!111! WHEN PATCH COME OUT!!!1!!1!!!!!!1111! PLS HELP!!!!!!!!1111oneoneoneone”, performance data that doesn’t represent a real world use case, etc), then Linus has given away his time without any contribution in return. So if the whippersnapper’s contribution *is* good, Linus praises it, the whippersnapper gains status and finds future projects he contributes to more likely to devote time to considering his work, and Linus himself gains status by being seen to praise good work. If it *isn’t* good, then he tears the whippersnapper a new one: he sacrifices status (Slashdot talks about how Linus is an asshole, film at 11) in order to damage the whippersnapper’s status pour encourager les autres (if you waste Linus’s time, he’ll let everyone know you wasted his time, which will make it harder for you to have future work considered by projects you submit code to).

    Of course, tearing someone a new one doesn’t *always* decrease Linus’s status. If everybody is already mad at a certain vendor, let’s make up a name and call them “NVidia”, and Linus flips them the bird, then the whole community says “Yeah, verb those adjective nounheads at NVidia” and Linus gains status.

    So the economy here is in time. Since the value that you will receive from allocating someone a slice of your time isn’t always obvious before you’ve allocated it to them, bolstering or attacking their status is used to adjust the value of the transaction after the exchange of time has occured (if they’ve wasted your time and the whole thing is a net loss for you, you make it a net loss for them too. If you gained by their use of your time, you reward them with status in addition to the time already given), and to signal to the community how likely it is that giving a given person a timeslice will be beneficial.

  8. TheDividualist, you’ve got it backwards: currency is basically a status/reputation. You put in an eight-hour day slinging hash or code and you raise your reputation, so that others give you the fruits of their own labor in return. What’s unique about currency is that it’s much more precise than other reputation systems: it allows you to trade the chickens you raised for the pigs they bred in proportion to how much time and material you spent on your work, how many others chose to compete with you to do the same work, and how much others want the results, ie cost, supply, and demand. However, at any time, the reputation currency we call dollars can be refused: take the bakers who don’t want to bake cakes for gay marriages (not casting judgement, I think that should be their right, however stupid I may find it). And of course, any status or reputation is useless on a deserted isle, same with currency.

    With the internet, there is a great opportunity to create much better reputation currencies. Bitcoin, Ethereum, and the other crypto-currencies are where your mind probably goes, but that is simply an attempt to decentralize something like existing fiat currencies. What’s coming are much more specific reputation currencies, like hacker or musician coins, that denote your particular wisdom or skill in specific niches. You can see the beginnings of this in systems like reddit karma or SO reputation points. Programmers already get hired based partially on their SO reputation number: that is the future for every profession and skill, whether fishermen or deciding who you want to play PS4 games with online.

    Regarding your note about Linus needing to lose reputation when he thanks someone, you’re thinking of reputation in analogy to the current dollar system of Fed Notes, where an external central bank like the Federal Reserve carefully controls how much currency there is. Whereas these new reputation systems will work more like free banking, you simply mint your own currency whenever you want. Since our smartphones can easily keep track of how much we’re minting, anyone can always normalize for how much two Linus bucks are worth, ie you ignore that he doesn’t have to lose his own status, because the value of the status conferred by his praise goes down exactly proportionally with how easily or often he does it.

    btw, esr, speaking of your past pieces, I wonder if you read this recent piece by a Microsoft Sinofskyite, pinning their OS market share decline on missing the rise of mobile computing and its constraints, rather than the 64-bit transition you were hoping would lead to a linux rise.

  9. @Nove

    There is another factor to watch regarding Microsoft: The rise of the Vulkan / AMD / Valve alliance.

    Vulkan is one of the new graphics APIs that are designed for the architecture of modern GPUs, instead of having that bolted on a-la OpenGL. Normally this wouldn’t matter that much in terms of market share; except that the new version of DirectX is sufficiently different from previous versions that developer knowledge does not carry over. If you have to learn a new API anyway, why not check out that Vulkan thinggy?

    It doesn’t take much reading between the lines to see that AMD (the primary backer of Vulkan) is directly positioning Vulkan as an openness play against NVIDIA. Add to this the fact that AMD has started pouring resources into the open source drivers.

    But wait! There’s more! Microsoft has been trying to close the garden again, this time with games. Such behavior is a direct threat to Valve’s business, so Valve is also pushing Linux more, including pouring resources into the open source AMD drivers.

    (oh, and shameless plug: Newtonian Money in the Ergosphere)

  10. Ian, I’ve heard of Vulkan, Mantle, and the like, but I don’t game so didn’t know about the renewed focus on OSS drivers.

    As for your linked blog post, I think you dismiss commodity backing too lightly, though I’m fully agreed on historical commodities like gold or cocoa beans. You’re right that when you’re usually trading digital tokens, it’s not worth backing it with a single, largely unused commodity like gold.

    But what is money? It is really a claim on future production, usually in the short-term, as most don’t hoard cash for longer periods these days. Currency was the first futures contract, useful to make an unspecified purchase sometime over the next couple months. Well, we actually have specific futures contracts these days, albeit only for certain commodities and usually only owned by companies or investors.

    What I see replacing money are futures contracts for the everyday goods we now buy with cash, ie you’ll have divisible futures contracts for everyday items like oranges or toothpaste. Rather than hold cash, you’ll hold a digital bucket of these futures in your smartphone, for all the household items you foresee buying in the next six months. When you go to the grocery store and check out with a dozen oranges, your smartphone will redeem your futures contracts for oranges or automatically execute a series of trades of some of your futures contracts for others that can be redeemed for a dozen oranges that day. It’s basically digital barter.

    What are the advantages? Well, for one, we get rid of the medium of exchange, no intermediate backing, whether e-gold or bitcoin, needed anymore. Another is that when you hold dollars or bitcoin, producers have no idea what you’re going to buy with it. Price is a signal, similarly these futures contracts you hold will signal your demand for oranges and your favorite brand of toothpaste. One month, you sell all your orange futures for apple futures, because you’re feeling more like the latter fruit, which, if seen in the aggregate too, signals orange producers to cut back on production.

    There are complications: it is harder to envision the relative value of goods, ie how many plums you could exchange for two apples, because you took away the current single numeric scale with which to measure, dollars. If you used to phone your friend to tell him you got a great smartphone for only $300, that will be harder to communicate when there is no single currency.

    But these are minor details, these futures contracts and digital barter are likely the endgame, though of course dumb regulators will try to fruitlessly slow it down. The problem with those pushing bitcoin and other online fiat currencies is that they don’t really think through what is newly possible online now, but simply recreate the existing solution, fiat currencies, in a more decentralized way. Much more is possible now.

  11. Nove, I can’t agree. The need to discover the relative value of goods efficiently is the principal function of money; it’s the feature that allows traders to factor complicated barter arrangements among many people into a series of pairwise exchanges. Using money as a store of value against unforeseen expenses comes after the money’s use as a medium of exchange is established.

    The retail futures scheme you envision could, to some extent, replace money as a store of value – insofar as people knew what their future needs would be, they could acquire futures contracts for things to supply those needs, instead of keeping cash until the needs arise. But they’d still need some way to store value in general, in case of need they couldn’t plan for. And the markets would still need a currency of account to let people know the relative values of goods they wanted to trade.

  12. Ian Bruene,

    You misrepresent the relationship of Vulkan, OpenGL, and the industry. OpenGL 4.x is designed for modern GPUs; what Vulkan gives you is efficient low-level access to GPU hardware. This is the same thing that DirectX 12 gives you and why DX12 is so different from DX11.

    Also, why would AMD position Vulkan as an openness play against NVIDIA? Before Vulkan, the only real players in the low-level GPU API space were AMD’s proprietary Mantle, Apple’s Metal, and DX12.

    DirectX won because unlike OpenGL throughout most of its history, DirectX’s API managed to not suck. I haven’t looked at Vulkan in-depth but given that it looks a lot less sucky and gets you high performance on Windows, Linux, and Android, that’s probably the reason why developers are taking a long hard look at it.

    Now, consider: iOS is still the premier mobile platform. Android wins in terms of sheer numbers, but in terms of number of apps and number of quality apps, iOS still leads. Virtually all professional mobile development is either iOS-only or iOS-first. iOS does not support Vulkan natively; Apple are committed to pushing Metal. Third-party wrappers exist, but these may not give you the performance that (heh) bare Metal does. With the bulk of gaming going mobile in a big way, is Vulkan’s future really so secure?

    And really, who gives a shit? Most game developers don’t interact with OpenGL, DX, Vulkan, or Metal directly; they use the APIs exposed by one of the Big Two engines (Unreal or Unity).

  13. A bit of an aside, feel free to ignore if you like – I’m very interested in General Semantics. Is there any particular source you’d recommend to someone trying to learn more about its methods and techniques?

    1. >A bit of an aside, feel free to ignore if you like – I’m very interested in General Semantics. Is there any particular source you’d recommend to someone trying to learn more about its methods and techniques?

      That’s a tough question. A lot of the sources I learned from are out of print. The really central source, Korzybski’s Science and Sanity, is (ironically) nigh-unreadable.

      Aha. I see there’s a free download of Language in Thought And Action available now. That’s a good starting place. There’s also Stuart Chase’s The Tyranny of Words, the earliest popularization.

      Ian Bruene correctly implies, though he does not outright state, that you can learn a lot of GS from studying Eliezer Yudkowsky’s Sequences. Yudkowsky has more inclusive goals than GS but was clearly strongly influenced by it. (Actually, it wouldn’t be crazy to suppose that some of that influence came through me.)

  14. M. Brazier, yes, currency was a great innovation over barter centuries ago, as it allowed precisely the simplification in calculation and coordination that you note. However, now that our smartphones and computerized point-of-sale terminals handle all the calculation, it is possible to go back to barter but in this more advanced digital form, so that we don’t need a currency anymore. Surely you’re not arguing that the computers can’t handle it for us.

    You wouldn’t use a “currency of account” to let people know the relative value, they’d choose a metric that worked best for them. Some may choose hours of work, ie rather than the current system that tells them the smartphone costs $300, and then they have to calculate that they would have to spend 20 hours at their $15/hour job to buy it, the software just shows them 20 hours as the price. The same for anything else they buy, whether socks or a car. Others may want to know what it costs as a percentage of their monthly income. It will vary by person, and this is only possible now that it’s all handled in software.

    Of course, you will need to have market makers who maintain software that keeps giant lists of relative valuation, just as any business or market maker maintains giant, fluctuating price lists in dollars now. Your smartphone would automatically interact with this market-making software to execute trades of futures to pull off purchases. But it’s not something the average consumer will have to think about, just as they have no idea how the Federal Reserve works or of the constantly fluctuating value of the dollar and retail prices now.

    Of course, there are a few leaks of abstraction. Communicating prices by voice might require a common index that everybody uses in that niche case, but that won’t be a currency. Emails could be automatically translated, ie you write your friend that you bought the smartphone for 20 hours of work, and that gets translated to the favored metric of the receiving person by software, eg 15% of their monthly income (the software would translate the emailed text “20 hours of my work” to the phrase “15% of your monthly income”). Some market-making software may choose to use an index or scale internally, rather than maintain relative valuations for every good, ie a vector price list rather a matrix of all possible swaps.

    But those are minor issues, the main idea still holds up. As for your objection of unplanned needs, simple, you’d keep extra futures of goods and services – oh yes, you will buy futures contracts good for one-tenth of a haircut at your local barbershop too :) – that you historically bought, and when you buy something not on that list, the smartphone simply trades them with market-making software for the futures you need.

  15. @ Jeff Read

    OpenGL 4.x is designed for modern GPUs; what Vulkan gives you is efficient low-level access to GPU hardware. This is the same thing that DirectX 12 gives you and why DX12 is so different from DX11.

    Ah, it has been a bit since I looked at the situation in detail. But the conclusion is nearly the same: the new APIs are better at letting the programmer get more out of a modern GPU architecture.

    Also, why would AMD position Vulkan as an openness play against NVIDIA? Before Vulkan, the only real players in the low-level GPU API space were AMD’s proprietary Mantle, Apple’s Metal, and DX12.

    Because AMD is pushing the open line in several different ways. And if Vulkan unseats DirectX suddenly the API that gets good performance on AMD GPUs is on top. Yeah, no advantage to AMD at all here.

    Labeling this particular part of AMD’s strategy an “openness play” is probably not quite accurate, I’ve also called it a “ubiquity play”, but that doesn’t quite fit either.

    DirectX won because unlike OpenGL throughout most of its history, DirectX’s API managed to not suck. […] Vulkan […] looks a lot less sucky […]

    True, True, and True.

    [iOS vs Android]

    Don’t care.

    With the bulk of gaming going mobile in a big way, is Vulkan’s future really so secure?

    Ok, I’m gonna need your definition of what counts as “gaming”. Because “Free to play”-bordering-on-scam Mobile Game #904762039473857349264* does not count. When “mobile gaming” includes games with the sophistication of let’s say…. the new Prey, then I’ll take this seriously. Ok, I’ll take lowest common denominator into account: when the newest Call of Duty Micheal Bay-fest comes out on mobile I’ll listen.

    * that number is probably a low-ball

    And really, who gives a shit? Most game developers don’t interact with OpenGL, DX, Vulkan, or Metal directly; they use the APIs exposed by one of the Big Two engines (Unreal or Unity).

    By this logic the current APIs are irrelevant. Pushing the decisions a step up the library chain doesn’t change the fact that those decisions get made. Besides, who do you think checks the box in the engine to use/support XYZ API in their game?

  16. @Davis

    You’d have to ask ESR, I pieced together my understanding of it from a combination of ESR’s writings, H. Beam Piper’s stories, Godel Escher Bach, and a sprinkling of lesswrong.

    @Nove

    And the utility functions of these expert-systems-bordering-on-strong-AI will still be using a unit of account, if only to save the sanity of the programmers. Just because you have cycles to burn doesn’t mean you should do things in the most complicated way possible. Saving cycles by using currencies will be useful for a number of reasons:

    * operating on ultra-low power devices (especially if you are talking about a proper implanted Muse where you really don’t want to cook the surrounding brain tissue)
    * spending cycles to find better trades
    * doing other stuff on the same hardware
    * not having your code being mocked on Daily WTF

    Having said that, yes, “personal currencies” and “company currencies” will probably emerge. The concept was all the rage with various anarchist thinkers back in the day, but they didn’t understand what a currency is for so they were speculating on something that couldn’t work with the tech they had available at the time.

  17. Nove, one could write a market-making software program that worked as you describe, but I doubt that anyone would want to. Your software would, necessarily, have to store and update exchange rates between all possible pairs of economic goods, just to function – requiring O(n^2) space. For any modern economy that’s impossible to manage; the number of available goods is in the millions.

    The only practical way to implement a market in software that scales to modern economies is an internal price index, giving the exchange rate between each good and an abstract unit of value; that’s an O(n) solution. But an abstract unit of value just is a currency of account. And if the software defines a currency of account, there’s no good reason not to show prices in that currency to the users. While it would be feasible for the software to translate each price to whatever unit a user prefers, including “time to afford at my daily labor rate” or “percentage of my monthly investment income”, I expect most people would use either the market’s currency or their local community’s money.

    IOW, a market implemented in software just separates the functions of “medium of exchange” and “store of value” that money usually combines. There’s nothing especially revolutionary about that; it’s been done before, without computers.

    As for unplanned needs: which goods would you hold a claim to, as provision against them? If we leave out the possibility that the market itself might not be available (the market’s software can’t deal with that case) the best choice is a good that anyone would value at any time, and which can be finely divided, so if a problem appears you need give up only as much value as you must to buy what you need.

    For this purpose, specialized labor is out – you can’t expect to sell 12.345 haircuts at a moment’s notice; any one person needs only one haircut at a time, and no one can use a fraction of a haircut. Labor future contracts aren’t much better. (I don’t think there even is a market for labor futures anywhere. Employers normally pay wages after receiving their workers’ labor.) Perishable goods are, well, perishable; they lose value if kept too long. And specialized equipment, or land, can’t be divided too far – no one can use half of a car, or one square foot of land. What does that leave? Stable chemicals with multiple uses … and money.

  18. Ian, you fail to take into account the only thing that truly matters in the gaming business: M-O-N-E-Y. The money to be made in free-to-play bordering-on-scam mobile games so dwarfs the money to be made in AAA that AAA publishers like Konami are pivoting to a mobile-centric business model. AAA gaming is turning into what comic books were in the 90s, a format and medium that caters to a niche audience. If AAA publishers are not making the most money, they don’t get to set the standards.

    Oh yeah, and as for Linux gaming? Like Linux on netbooks, it’s a mere bargaining chip. By threatening to move Steam to Linux, Gaben can get Microsoft to play nice. Once Microsoft starts playing nice, the Linux-gaming effort will be abandoned as Valve renews its ties with Microsoft on much more favorable terms.

    So far, this seems to be working: Microsoft is even committed to releasing game content, like Forza 6 and Killer Instinct, on Steam. Between this and Steam Machines being an absolute bust, I do not expect Linux to play a significant role in the gaming scene in the future. Which negates some of Vulkan’s value proposition (though not all: last I heard, Vulkan was much more performant than D3D12 on the same hardware with the same OS).

  19. Oh yeah, and as for Linux gaming? Like Linux on netbooks, it’s a mere bargaining chip. By threatening to move Steam to Linux, Gaben can get Microsoft to play nice.

    This is news to approximately no one.

    Once Microsoft starts playing nice, the Linux-gaming effort will be abandoned as Valve renews its ties with Microsoft on much more favorable terms.

    At which point all of Valve’s contributions will vanish in a puff of logic.

  20. > With the bulk of gaming going mobile in a big way, is Vulkan’s future really so secure?

    Mobile has interface issues, and as a result, the war for mobile isn’t won yet, so no 3D API is yet secure. There are certain types of games (and programs on general) that can’t really be done without a mouse and keyboard. And the interface styles needed for mobile vs. desktop apps and DEs to be usable are generally disjoint. That’s why pre-iPhone attempts at mobile computing didn’t really take off: they generally tried to use a desktop interface style with touch or a stylus, which resulted in a system that wasn’t really usable on the move. It’s also why Windows 8 and various dockable Android schemes suck: They attempt to use a mobile interface style when docked, which makes them useless for anthing that requires a mouse and keyboard. Microsoft has gotten their head at least partway out of their rear with Win10, but still has to deal with the fact that most of their existing Desktop apps are closed-source x86, while the most battery-friendly CPUs tend to be ARM. Canonical’s Ubuntu Phone boondoggle went south for the same reason: they were originally on the right track with Ubuntu for Android, but then got sidetracked by the idea of “convergence” and put UfA on the shelf and started working on a system that would try to be as consistent as possible across its mobile and desktop interfaces, which is where every other desktop-on-a-mobile-device project has sunk: as soon as you try to use the same interface docked and undocked, your device is useless in one or both of the two situations.

    My prediction is that the platform that will win the war for mobile will be the first one that can offer a pocket-sized phone (not a tablet) that, out of the box:

    1) Is dockable
    2) Runs a decent desktop environment when docked.
    3) Can run most existing desktop apps for one of Windows, Mac, or Linux.
    4) Presents a decent mobile interface (not a clone of its docked interface) when undocked.
    5) Possibly less critical, but will certainly be very helpful, can run mobile apps for some existing platform, the more market share, the better.

    MS could still win if they can release a phone that presents the Win10 desktop when docked and can run existing x86 apps well before anybody else releases a device that meets the criteria above.

    Linux is a strong contender: Most of its existing desktop app base is open source, so it doesn’t run in to the binary compatibility issues that Windows does, and it’s at the core of Android. If some vendor releases an AOSP-based phone that runs a bog-standard Linux DE when docked, Linux will almost certainly become dominant everywhere that it isn’t already, by every metric from market share to developer revenue share.

    It’s unlikely, however, that Google will include this capability in stock Android, or that Apple will release a device that meets my criteria: both companies have business models that would discourage them from releasing such a platform, even if failure to do so would be shooting themselves in the foot in the long term.

    For Google, it’s that they don’t want end-user devices to really be all that capable: they want them to be terminals to Google services, and Android devices that could run desktop (and server! ) Linux software would be counter-productive to that goal.

    For Apple, it’s that they’re in the business of selling hardware, and an iPhone that could run MacOS docked would cannibalize the market for Macs. They wouldn’t be able to sell iDolaters an iPhone *and* a MacBook anymore, they’d just be able to sell them an iPhone. There also might be issues with binary compatibility, but given that Apple has experience with doing wholesale architecture switches, I doubt that transitioning desktop MacOS from x86 to ARM would be nearly the problem for them that the equivalent transition would be for Microsoft.

    Now, both Google and Apple might wake up and realize that they could still lose if someone else manages to bring the desktop to mobile first, so either iOS or stock Android might still win. And even if Google doesn’t do it with stock Android themselves, they might allow some Android vendor to include a full GNU/X userland on a device (although some of the anticompetitive contracts they have signed with vendors make me less than confident that they would), and even if whatever vendor eventually brings the desktop to mobile uses an AOSP fork without the Google Play infrastructure, Google could still adapt to and possibly even profit from the situation. But it’s still very much up in the air who will eventually win. My guess at the most probable winners is (though I’m not entirely sure of which order 2 through 4 should go in):

    1) An AOSP fork running a Linux desktop.
    2) An “Ubuntu Phone done right” project, running an existing Linux DE and desktop software stack alongside a new mobile environment.
    3) A third party vendor bolting a Linux DE onto stock Android.
    4) Windows.
    5) Google integrating a Linux desktop into stock Android.
    6) A “MacPhone” from Apple.

    For a while it looked like UfA might happen as either 1), 3), or 5). UfA might still happen as 5) if Canonical and Google stop playing reverse Ouroboros and partner up, but as they’re about the two companies in the Linux community that are worst about playing well with others, I wouldn’t count on it. For a while it looked like Ubuntu Phone might happen as 2), but the project went off into the weeds and is now vaporware.

    If MS wins, they will probably dominate for another 20-30 years, though due to the weakness of their existing mobile platform, the other contenders will likely survive. If Linux or Apple wins, MS will likely collapse, though they might survive if Linux wins with a mobile platform developed from scratch rather than Android. If Linux wins via Android, Apple will be in trouble, though it likely won’t kill them. If Linux wins via a new mobile platform, Apple will be able to maintain quite a strong position if it acts quickly, and might even be able to regain dominant market share if its desktop-on-mobile option can eat Android’s market share in the mobile space and Windows’ market share in the desktop space before the Linux platform can get momentum going on the mobile side. If Apple wins, Google will probably survive and introduce a GNU/X userland on Android, but Apple will dominate computing for the next generation, and non-Google Linux options will collapse in the desktop/mobile space.

  21. MS could still win if they can release a phone that presents the Win10 desktop when docked and can run existing x86 apps well before anybody else releases a device that meets the criteria above.

    It might even happen this year. Microsoft already has the emulation working on ARM parts. The only thing holding it back is IP issues. Intel holds patents to much of the x86 ISA; anyone distributing emulation of modern x86 or x86-64 architecture will have to license up with Intel first — which fees could run nine figures easily.

    And yes, the Intel patents cover the instructions themselves, irrespective of implementation. Intel has successfully defended this IP in court against the likes of Transmeta. Don’t like it? Write your Congressman. Or move to a more innovation-friendly venue like Western Europe. Or China.

  22. It’s interesting: Intel doesn’t have a lot of good options here:

    If they’re too aggressive with their patent licensing fees, then they make it less likely that the first desktop-on-mobile platform will be from MS or Apple, which are the OS vendors most dependent on the x86 ISA, due to the amount of x86-only software for their platforms with no publicly available source.

    If they’re not aggressive enough, Apple and MS become less dependent on x86.

  23. “MS could still win if they can release a phone that presents the Win10 desktop when docked and can run existing x86 apps well before anybody else releases a device that meets the criteria above.”

    The outcome depends on the relative costs of bandwidth and cloud computing. If the costs of bandwidth and cloud computing are low enough, it is more efficient to use thin clients than managing a full software installation. Then Google probably has the better cards with chromebook like products. If not, MS might make a comeback with fully capable computer terminals.

    But the history of Windows illustrates how extremely difficult it is to make inroads into an established market. And for all purposes, Android has the mobile space covered with iOS as a distant second. And Apple never showed any desire to carter to the poorer 80% of the market.

    Windows has already shown to be able to utterly destroy even the strongest most dominant brand in mobile, Nokia. So I would not hold my breath for anyone in mobile to bet on Windows again.

  24. Ian, I already noted above that some market-making software might just use an “index or scale,” but I’m open to that not being necessary. The point is that it will be up to the market makers to decide how they internally implement it, and even though you and I can’t see why they wouldn’t use an index, perhaps not using one will prove to be useful in some way. The client will simply put up an auction of one type of future for another that the user needs and all the market-makers will respond with their best bids. How the market-makers eventually end up implementing their systems internally is somewhat irrelevant to the discussion but I don’t think efficiency will be a big deal for maintaining relative swaps, as it will all be done in the cloud and segmented by market, ie market-makers will specialize in products like tech or textiles and federate when needed.

    As for personal and company currencies, I’m actually talking about getting rid of currencies and referenced neither. I talked about a separate concept of reputation currencies initially, ie something like hacker or musician coins, but that’s fairly different from this futures concept. I did mention “Linus bucks” in that context, but that was a flippant remark based on the fact that he’s so famous in tech circles that he could have his own reputation currency. :) It’s more likely that he’ll simply have a lot of generic hacker coins, instead of his own Linus bucks, and be able to spread the wealth around by paying for good work with those new coins, along with the thanks he gives now.

    M. Brazier, as noted above, with current hardware capabilities and federation, I don’t think handling all possible pairs would be a big deal. I don’t think that’s necessarily the way it will evolve, but I’m open to seeing how it works out, ie market makers will experiment and figure out the better way. But let’s assume that many market-makers internally end up using some vector “price” list keyed around some “unit of value.” The point is that each market-maker will likely end up using their own preferred index or scale, one could use a Big Mac index while another uses some kind of CPI-derived unit, and a thousand arbitrary internal-only scales certainly wouldn’t amount to a “currency of account.”

    And I completely disagree that users would choose a currency or common index over the way they actually think, which is in terms of hours or work or percentage of income. Every person has their own internal scale for figuring out whether something is worth buying- for you, it may be how many Big Macs you could buy with the same amount of money, for me, it may be what percentage of my discretionary income it bites into- and the software will now be able to show the “price” to them in that idiosyncratic scale, without requiring further calculation on their part. That’s a huge win for users, one you discount way too lightly.

    I do think there will likely need to be a common regional index at some point, purely for legacy purposes. People who communicate prices in person or over the phone will need it, as I mentioned, and there will always be old people who need a “price” to replace the dollar price they’re used to. But that will be some sort of artifical index put out by several large market-makers, it won’t be a currency.

    Since there’s no “medium of exchange” in the system I’ve sketched out, I don’t see how you think this “market implemented in software” keeps it. I’d love to hear how you think something like this was done without computers, as it’s computers and the internet that makes this all possible now.

    Regarding unplanned needs, you’re taking this in a completely different direction, as you originally talked only about goods that you didn’t happen to plan for ahead of time, whereas now you appear to be talking about disaster scenarios? I already noted that simply holding extra futures of what you normally buy is enough, you’ll simply trade for the unplanned futures you need at that moment.

    If all the software itself is somehow down, guess what? Much of the market is already unavailable, even if you happen to be holding antiquated paper currency. And nobody is taking your “stable chemical” or gold at Walmart. Your notion of needing something that holds value “at any time” and can be finely divided is hilariously antiquated. Nobody has used such “currencies” in more than a century and even fiat money like dollar bills are about to be replaced.

    And yes, there will be divisible futures for all the possibilities you list and they will always be tradable. Even if no other end user wants them at any given moment, there will always be market-makers who offer a bid on them at some level they deem worthwhile and hold them till they’re more valuable again. No one can use a fraction of a haircut, but they will buy fractions because they may not be sure if they want something yet, but want to lock in the lower futures price in case they do, or because they can only afford a fraction right now, ie they slowly buy fractional futures of that car over time. As time goes on, they either purchase when their fractions make up a whole or sell the fraction they have when they realize they don’t want it.

    Jon, I wonder why you think the “war for mobile” isn’t already over, like everybody else. But no, adding a desktop interface option won’t make a difference to the mobile players, as you fail to notice that there are a lot more people who want the mobile interface than a desktop and they’re already well-served by the current winners, Android and iOS.

    However, the recently released Samsung Dex for the Galaxy S8 does most everything you want, with the possible exception of 3), though it has Photoshop and Office. I don’t know why you think Wintel apps still matter, in this time when everybody is rushing to get on mobile Android and iOS, including MS, Adobe, and all the former Wintel app vendors. Google doesn’t really care which platform you’re using as long as you’re searching with their engine. The multiwindow functionality in Dex was written by and baked into Android 7.0 Nougat by google. Apple doesn’t really care about Macs, they’re trying to cannibalize their own desktop interface with the iPad Pro. Tim Cook always talks about using an iPad himself, not a Mac.

  25. I did mention “Linus bucks” in that context, but that was a flippant remark based on the fact that he’s so famous in tech circles that he could have his own reputation currency. :)

    If reputation currency among hackers were a real currency, one of the coins or bills might have Linus’s picture on it :)

  26. Nove: A matrix of barter exchange rates is a quadratic algorithm for running a market. Introducing a currency is a linear algorithm for the same purpose. No sane programmer will use a quadratic algorithm when a simple linear algorithm is available.

    For example, a list of prices of 2^20 products, with 4 bytes for each price, takes up 4 Mb of storage. Keeping the same data as an all-pairs matrix, with 4 bytes for each exchange rate, would take 4 Tb of storage. It multiplies the storage requirements by more than a million. No one would choose that in production-grade software. All serious markets implemented in software would use an internal currency, for that reason alone.

    And I think you’ve been talking to some rather odd people. Everyone I’ve ever met thinks of their income and expenses in terms of money, not as fractions of their working time or any other index. It’s genuinely easier to think that way, and not just because prices are published in money. Consider people who get income in multiple forms – salary, commissions, and investments, say (an experienced salesman could easily earn all three at once.) How do they figure out how much they can afford to buy, if they don’t convert all their income streams to money?

    Regarding futures contracts: Let’s suppose that you’ve bought futures for all the haircuts a barber performs over the next two months, and then the barber breaks his leg and is confined to bed for that long. Now your futures in haircuts are worthless; the barber physically can’t provide the service that you theoretically have a claim to. Buying the barber’s labor futures has transferred his risk to you – making them the opposite of a store of value. The point of a store of value, remember, is to hedge against unforeseeable events.

    The idea of buying fractions of a good over time, instead of the whole price up front, isn’t a futures contract: it’s called “layaway” or “an installment plan”. It’s not done as much now, in the days of credit cards, but it used to be routine for things like furniture and appliances.

    On desktop UIs for smartphones: have you ever tried to write a long document on a phone? Don’t. A touchscreen UI can just about manage composing a Twitter message of 140 characters. For anything longer than that you need a proper keyboard. If all you need to do on a computer is read documents, the mobile UIs are enough for you. But for typical office work, or engineering? Only a desktop UI can serve your needs.

  27. M. Brazier, 4 TB is nothing when federated across tens of thousands of market-makers in the cloud, each of whom focuses on some niche of the market that they know well and thus splits up the matrix, surely you know that. I don’t know why you keep banging on about how the market-makers might someday implement their software internally, when all that matters is that they will provide bids on swapping one fractional futures contract for another. To make that possible, those market-makers might internally maintain giant matrices of swaps, flatten them out to vector “price” lists of some internal index unit, or standardize with other regional market-makers on some external agreed-upon index unit, but at no point will a currency be involved.

    People don’t think in terms of money, because for most people, money is not what they live their life by. They think about whether they’ll able to make their rent or a down payment on a car, and whether upgrading their smartphone early will make them miss either goal. In other words, they internally trade off on what purchases, or occasionally savings, they’re rather make, the money itself is meaningless. For those paid on a flat hourly basis, say $15/hour, buying that $300 dishwasher means 20 hours of work. That’s a very concrete thing for them, so they internally translate into that metric.

    I suggest you read up on how people actually think and calculate, rather than the rare finance types you mention. Your question is meaningless as stated: such finance professionals are currently paid in money, so there’s nothing to convert. But the question isn’t how they figure out how much they have, but how they decide to make purchases, as that’s what we’re discussing, and you’ll find that many even in finance would rather see prices in something other than currency. If you mean, how would finance professionals get paid when there’s no currency? Just like anyone else, with the futures contracts and longer-term equity and debt that they choose. :)

    As for your barber example, what do you do now? Either you love your barber so much that you don’t go for two months, or someone takes his place and you just get your hair cut by his replacement. Behind the scenes, your futures contracts will either be pushed back a couple months by the barber or swapped for those of his replacement, depending on which you chose and what he arranged for, but the actual futures transactions would be completely opaque for the vast majority of users. They will simply choose to wait or pick the replacement, and that choice on the barbershop’s online scheduler will trigger all the necessary futures transactions on their smartphone to make it happen.

    It is possible that some niche futures- let’s say hardware kickstarters ;) – don’t have such easy mechanisms in place, and then your fractional futures contract, worth a couple cents in today’s dollars, would go to zero. Oh well, you lose or gain a couple cents every time the current dollar or dollar-denominated goods and services you want to buy fluctuate in value every day: life will be no different. In other words, every “store of value” fluctuates constantly, whether dollars, gold, or these futures contracts, nothing changes in that regard.

    Yes, layaway plans have been replaced with credit cards and other types of easier debt, and they will be replaced in turn by fractional futures. The fundamental concept is the same, but they vary somewhat on when you take delivery of the item and consequently, how the interest rates are structured. Because of these details, they’re not identical and this newer construct allows more flexibility and power, just like many new financial innovations like credit cards have in the past.

    I don’t know why you stress the need for a “desktop UI” to get work done, when I noted that such a smartphone already exists, the Samsung Galaxy S8 and its Dex dock. If you’re rebutting my initial claim that the desktop interface won’t make much of an impact on mobile sales, perhaps you’re unaware that the mobile market dwarfs the PC market. I agree with you that you need a multiwindow interface to get real work done- I even provided you with an example of a popular smartphone that provides it- but if you think that niche use, productivity apps, makes much of a difference in the gargantuan mobile market, you haven’t been paying attention. ;)

  28. Nove: “4 TB is nothing when federated across tens of thousands of market-makers in the cloud, each of whom focuses on some niche of the market that they know well and thus splits up the matrix, surely you know that.”

    How do you federate the markets for, say, beets and tractors, without a currency? Beet farmers need to sell beets and buy tractors. If those are traded in unconnected markets, farmers can’t transfer what they get for a beet crop to the people who sell tractors, which means they can’t buy a tractor when they need one.

    Once you start tracing out all the factors that go into producing any single good, and all the purposes to which a single good can be applied, you’ll soon realize that there is no logical way of splitting the matrix of barter exchanges. Every good must, in principle, be exchangeable for every other good. The best you can do is what the inventors of currency have already done: factor the matrix, store it as the outer product of one row of prices with itself.

    “Behind the scenes, your futures contracts will either be pushed back a couple months by the barber”

    Ah, no. You can’t push a futures contract back – by definition, futures contracts are promises to deliver something on a specific future date. And you can’t trade an appointment with Bob the barber on Saturday for anything else once it becomes known that Bob can’t cut hair on Saturday – not even close substitutes, like a haircut from Bill the barber on Saturday, or a haircut from Bob next week when he’ll be better.

    You’re pushing the concept of futures contracts farther than it can really go. Futures don’t work well in any of the functions traditionally served by money. They’re no better than any other good or service as a medium of exchange, and definitely worse than the traditional forms of currency. And they’re singularly bad as stores of value – I think I could prove that a futures contract is necessarily worse, for that purpose, than the underlying good would be.

    “I suggest you read up on how people actually think and calculate, rather than the rare finance types you mention.”

    I said nothing about professional financiers. My example was “an expert salesman”. Salespeople, as you know, are people whose basic skill is persuading others to buy things; they are not found only in the financial markets. And most of them work on commission, in part or wholly. Don’t take hourly employees as typical.

    And no, I don’t believe that most people who work for hourly wages calculate their budgets in terms of work-hours, instead of money, even though they could. Still less do I think that people with more complicated sources of income would even be able to figure their budgets without using money.

    “I don’t know why you stress the need for a “desktop UI” to get work done, when I noted that such a smartphone already exists, the Samsung Galaxy S8 and its Dex dock.”

    The Samsung DeX is missing at least one of Jon Brase’s conditions: it doesn’t run most applications that were written for one of Windows, Mac or Linux desktops. It runs a subset of apps written specifically for Android 7.0. When Android 7 runs GNOME or KDE out of the box, that’s when mobile hardware will seriously contend with the Wintel platform; but we’re not there yet. Samsung DeX just shows it can be done.

  29. I suggest you read up on how people actually think and calculate, rather than the rare finance types you mention.

    The only people I’ve ever heard of thinking the way you claim most people think are economics geeks (and I count myself as one). And even then they only do it for examples.

    But ignoring everything else still leaves us with the 64 quintillion dollar question: What on earth is this system getting you that couldn’t be done with greater simplicity for everyone involved by just having a currency + Mk1 futures contracts / pre-scheduling + a money management app.

    This is the problem with all of the Up-end Economics For Great Justice! plans: the useful parts can always be easily simulated on top of a regular economy. And doing it that way results in far greater robustness and simplicity, not to mention ability to scale.

  30. @Michael Brazier:
    >When Android 7 runs GNOME or KDE out of the box, that’s when mobile hardware will seriously contend with the Wintel platform;

    Strictly speaking, mobile *hardware* already does contend with Wintel. What’s missing is a a good DE with an existing base of desktop-focused applications. Depending on the quality of DeX as a DE, it’s conceivable that apps developed specifically for it (as opposed to touchscreen usage) might accumulate and it might become the “winning” platform, but it would be starting from scratch and any desktop-on-mobile setup introduced in the meantime that *is* able to run applications from an existing desktop system would steal its thunder. Also, given the fate of similar systems (it’s not the first time that someone has tried putting together a DE to run Android apps on a desktop when docked), I think it’s fairly unlikely to take off without the help of an established app base.

    @Nove:
    >If you’re rebutting my initial claim that the desktop interface won’t make much of an impact on mobile sales, perhaps you’re unaware that the mobile market dwarfs the PC market.

    As soon as there’s a decent desktop-on-mobile platform, the mobile market won’t just dwarf the PC market: it will entirely absorb the laptop market and eat most of the remaining desktop market. I doubt that there are many people that use a mobile device but don’t use a PC at all, more probably, households that have individual mobile devices are sharing PCs. Once the desktop versions of programs like MS Office / OpenOffice can be run on mobile, there will be no reason other than performance (which won’t be relevant for most people) to have a PC at all. Sure, mobile versions of these programs exist, but on touchscreens they’ll be crippled by the limitations of touchscreens, and the design choices needed to make them usable at all on touchscreens will hobble them to a lesser degree on the desktop.

  31. Once the desktop versions of programs like MS Office / OpenOffice can be run on mobile, there will be no reason other than performance (which won’t be relevant for most people) to have a PC at all.

    If that’s the case then say goodbye to any sort of open hardware which can run any OS. Your only choice will be to run the kernel signed and supplied by the vendor; and you will be at the behest of the vendor when it comes to upgrades — as is already the case for all Apple and Windows-based mobile devices, and most running Android.

    I am, of course, speaking of the consumer market. Workstations that allow end-user OS installs will still be available — they will just cost five to ten times as much. Kinna like how you can still buy floppy disks or LCD panels with a 4:3 aspect ratio, but you have to pay the premium associated with such niche, boutique products.

    1. >If that’s the case then say goodbye to any sort of open hardware which can run any OS.

      This will not happen for the exact same reason it has failed the last several times it has been tried. The data-center and corporate customers – who are the real volume markets for white-box PC hardware – won’t stand for it. They see no reason they should pay “five to ten times as much” to preserve deployment flexibility just because Microsoft’s business model needs propping up, and they’ll stiff-arm any vendor who plays along with the lockdown.

      Apparently you learned nothing from the failure of Palladium. The white-box vendors got the lesson, though, and can’t afford to lose those sales; therefore, they’ll tell Microsoft to fuck off politely before the data-center people have to do it rudely.

      1. @esr

        You have failed to embrace the Wisdom of Jeff Read. The biggest, nastiest bully always wins.

        No Exceptions.

        1. It’s not just about bullying. It’s about market forces. Bully or not, companies will do whatever the market will let them get away with (absent government regulation, which is why we have government regulation). And the market didn’t bat an eyelash when Apple sold locked-down, walled-garden devices because those devices still enabled things which were previously not possible.

          Locking the device down may even confer an advantage in the marketplace. Nintendo’s locked-down console absolutely trounced all other comers in the marketplace since it allowed Nintendo to prevent third-party software from tarnishing their brand and associate their brand with quality (even if the “Seal of Quality” was sometimes applied to crappy games). But beyond that, whitelisting is still the best way to provide some semblance of security assurance in a world of billions of networked universal Turing machines. Since then, lockdown has been the standard for game consoles.

          Actually come to think of it, iOS is not only the most secure mobile OS, it’s quite possibly the most secure end-user-oriented OS ever written. Apple has really thought through the security thing, right down to the silicon; they even isolate the baseband chip from the main CPU in a way that many Android devices do not. Unfortunately part of the package when you get a device that offers the strong security assurances an iOS device does, is you don’t get to run another OS or unauthorized software. Jailbreaking is dead; if a vulnerability is found in recent iOS versions that enables jailbreaking, then it is much more valuable on the black market where it can be sold to criminals and oppressive governments for millions of dollars. The fact that rooting Android devices is still a thing is only testament to the fact that compromising Android is a doddle compared to iOS.

          (Scary thought: There’s a 20-year-old kid out there who’s richer than you because the Ayatollah — or the CIA — wants to find and murder their iPhone-carrying adversaries.)

          So, which would the mythical Average User rather have — a PC, with its open architecture and concomitant vulnerability to malware, OS- or sub-OS-level rootkits, etc., or a device with the functionality of a PC but the security profile — and App Store experience — of an iPhone? The answer to this question will determine which sort of device will benefit most from economies of scale. This won’t preclude other such devices from being available, but they will be rarer, more expensive, and maybe only available to large corporate customers buying in volume.

          1. >It’s not just about bullying. It’s about market forces.

            Market forces, eh? So, how’s that “peak oil” prediction working out for ya?

            Jeff, when you get enough of a fucking clue about demand elasticity or substitution to not make a fool of yourself calling a market as simple as energy, maybe I’ll listen to you honk about “market forces” without laughing my ass off.

      2. Talking about the failure of Palladium is like talking about the failure of the U.S. government’s Total Information Awareness initiative, or Scientology’s Fair Game policy. These things didn’t fail as such; they were publicly dismantled to avoid continuing PR backlash, but the pieces were implemented quietly under different names. Most motherboards contain a TPM now, and Windows uses it to encrypt on-disk content (BitLocker). The trusted-boot-path bits were more or less implemented in the form of UEFI Secure Boot, and these days hardware manufacturers are under no obligation to allow disabling of Secure Boot or installation of user-provided keys. The main bit that seems to be missing is the “curtained memory” that allows scary things like DRM modules running outside of supervision even by the operating system; this could be achieved with a hypervisor and separate OS images to isolate the DRM modules, as is done currently on the Xbox One operating system with its separate, concurrently running OS instances for games, the Xbox Dashboard, and Windows 10 Universal Apps. Apple’s Secure Enclave is another version of the idea.

        The ideas behind Palladium are still with us, and to be honest, they’re not all bad. It’s just very easy and tempting to use them for bad things. I sincerely doubt that the corporate buyers of white-box hardware will be ill affected by Microsoft tightening the noose on end-user devices. They are already paying a premium for their enterprise-grade hardware, and if you’re looking to outfit a datacenter you’re probably in a good position to negotiate volume discounts to partially offset that premium anyway. It’s the end users — the pool from which future hackers come — who will be screwed.

        1. > Most motherboards contain a TPM now, and Windows uses it to encrypt on-disk content (BitLocker).

          Windows can do whatever the fuck it likes without rescuing your argument. We can still boot Linux and *BSD, and because of datacenter and cloud customers that will not change. All the squid ink you’re now squirting into the water cannot obscure the fact that Microsoft failed at this lockdown when it was at the apex of its monopoly power. Microsoft is not going to succeed now, not when it’s fighting a rearguard action to even stay relevant in that market by implementing WSL.

  32. M. Brazier,

    >How do you federate the markets for, say, beets and tractors, without a currency?

    Are you familiar with the concept of barter? :D There will be agricultural market-makers who focus on futures of beets and carrots and industrial market-makers who focus on tractors and backhoes. There will be intermediate market-makers who interface with the agricultural market-makers for both common farm tools, like spades or shears, and produce like beets and carrots. Put together this long chain of digital barter and you’ll be able to go from beets futures to tractors, all within milliseconds. :)

    >Once you start tracing out all the factors that go into producing any single good, and all the
    >purposes to which a single good can be applied, you’ll soon realize that there is no logical way of
    >splitting the matrix of barter exchanges. Every good must, in principle, be exchangeable for every
    >other good. The best you can do is what the inventors of currency have already done: factor the
    >matrix, store it as the outer product of one row of prices with itself.

    It is fascinating that you believe that simply introducing a currency magically solves the intrinsic circularity of determining the relative valuations of all goods. :) It doesn’t. All it does is provide a single dollar-denominated price vector to replace the barter matrix, one that will be replaced by the federated matrix I’ve sketched out above.

    >Ah, no. You can’t push a futures contract back – by definition, futures contracts are promises to
    >deliver something on a specific future date. And you can’t trade an appointment with Bob the
    >barber on Saturday for anything else once it becomes known that Bob can’t cut hair on Saturday
    >– not even close substitutes, like a haircut from Bill the barber on Saturday, or a haircut from Bob
    >next week when he’ll be better.

    It is a contract, you can make it do anything you want. You simply put in an electronic clause that says that Bob can push back the date on his personal services futures contract if he needs to call in sick, or provide you with a replacement futures contract from Bill. You aren’t trading the futures contract when Bill shows up, it’s a replacement provided by Bob.

    >You’re pushing the concept of futures contracts farther than it can really go. Futures don’t work
    >well in any of the functions traditionally served by money. They’re no better than any other good
    >or service as a medium of exchange, and definitely worse than the traditional forms of currency.
    >And they’re singularly bad as stores of value – I think I could prove that a futures contract is
    >necessarily worse, for that purpose, than the underlying good would be.

    Yes, keep repeating your mantra, “Money is great, nothing can replace money.” :) Money was the first crude futures contract: it’s about to be replaced by actual futures contracts that are much more flexible and powerful.

    >I said nothing about professional financiers. My example was “an expert salesman”. Salespeople,
    >as you know, are people whose basic skill is persuading others to buy things; they are not found
    >only in the financial markets. And most of them work on commission, in part or wholly. Don’t take
    >hourly employees as typical.

    I know of no non-finance salesman who makes a significant share of their income through “investments.”

    >And no, I don’t believe that most people who work for hourly wages calculate their budgets in
    >terms of work-hours, instead of money, even though they could. Still less do I think that people
    >with more complicated sources of income would even be able to figure their budgets without
    >using money.

    Nobody will “calculate their budgets” soon, it will all be done in software. We’re not talking about monthly budgeting but individual purchasing decisions: what comparisons, trade-offs, and calculations are made in a person’s head when deciding whether the “price” of a good is worth paying. You provide no theory of how those decisions are made, but simply disagree that they weigh the decision in hours worked or as a percentage of their income. It must be great to demur and simply invoke magic fairy dust as the reason.

    As for “complicated sources of income,” I’ve already noted that the sources are irrelevant. It is trivial to add up dollar streams now, it will still be trivial for your smartphone to do so when you’re paid for your various income streams with the futures contracts and debt and equity that you prefer.

    >The Samsung DeX is missing at least one of Jon Brase’s conditions: it doesn’t run most
    >applications that were written for one of Windows, Mac or Linux desktops. It runs a subset of
    >apps written specifically for Android 7.0. When Android 7 runs GNOME or KDE out of the box,
    >that’s when mobile hardware will seriously contend with the Wintel platform; but we’re not there
    >yet. Samsung DeX just shows it can be done.

    Yes, I noted that his condition 3) was only partially fulfilled when I first mentioned the Samsung Dex. I don’t know why you feel the need to repeat my point. And no, Dex doesn’t run apps written just for Android 7.0, it runs all Android apps. Lol, GNOME or KDE, what a joke, the Android productivity experience today is likely leagues ahead of either of those.

    Nobody cares about “seriously contending with the Wintel platform.” Did you look at the chart I linked above? The computing market is the mobile market, while the Wintel PC is a legacy computing platform, like mainframes or Unix workstations. When mobile reaches saturation, they will spend some small portion of that massive scale and simply swallow up most of the productivity market too, as Jon notes. But right now, they don’t have to care because practically all personal computing device sales is of the mobile platforms of Android and iOS.

    Your arguments have been hilariously loopy, so I’ll leave it here with you.

  33. Ian, how is it exactly that you claim to know how people think about how they make purchases? I’m basing it on how I think and some articles I read. I think the majority of people actually have a lot of trouble dealing with dollars, that’s a big part of why they usually make poor purchasing and financial decisions. I believe they will be able to make better decisions when presented with these metrics that better reflect the way they think and their costs.

    As for your big question, you get rid of currency because it’s an outmoded concept and because it distorts the market. Historically, the costs of barter were so high that it was worth setting one commodity, like gold or seashells, as the medium of exchange, despite the fact that that new use of gold as a money, in addition to all its other uses, would raise the price of gold, because of its new demand as money.

    People didn’t gold rush because they needed dental fillings or jewelry, but because there was literally money in the ground, which when brought back en masse would cause inflation. That brings us to our current currency, fiat money, which can and has been inflated by despots all over the world. Currency is fundamentally an information good: it conveys how much work you’ve done and how valuable it was, measure of value, keeps that information for the medium-term, store of value, and provides a way to coordinate purchases, medium of exchange.

    Well, we now have tech that does all that information work for much less than the costs of keeping the antiquated currencies around, by going back to digital barter. Not only will it be much more efficient, but we get rid of all those distortionary effects of currency, which you have internalized so much that you do not even realize them anymore. This scheme has nothing to do with “Great Justice!,” it’s simply more efficient and removes the distortions of currency. I see little difference in the scalability, robustness, and simplicity when compared to digital currencies- there’s no contest if you’re comparing digital barter to physical currencies, the former wins in a rout- merely that you are used to the old solution and unused to the new one.

    Jon, you overestimate the value of existing desktop apps, everybody wants to be on mobile first these days. As for prior attempts at running Android apps in a multiwindow interface, they were all fairly botched experimental attempts by nobodies. This is a device that will sell in the tens of millions by the second-biggest computing hardware company in the world, a completely different league.

    I agree that the mobile market will eat most of the PC market soon enough, but there are a lot of people in developing markets for whom their smartphone was the first “computer” they ever touched. They’d never had a PC or internet access before their smartphone, I’ve met these people.
    If they need to run some productivity apps some day, they will use products like this with their smartphone, the most funded hardware kickstarter ever.

    Anyway, my point was that whenever the mobile market goes after that small legacy niche, it will make little difference in who wins the mobile wars, because as the chart I linked shows, there’s just not much volume in that shrinking PC market. You initally claimed that this small legacy PC niche would determine who wins mobile, when the fact is that a lot more people want a mobile interface than a desktop interface. So the tail cannot wag the dog, and soon the dog will simply eat its tail. ;)

  34. Nove: “There will be agricultural market-makers who focus on futures of beets and carrots and industrial market-makers who focus on tractors and backhoes. There will be intermediate market-makers who interface with the agricultural market-makers for both common farm tools, like spades or shears, and produce like beets and carrots. Put together this long chain of digital barter and you’ll be able to go from beets futures to tractors, all within milliseconds.”

    So you envision a massive graph of markets which links each good only to its factors of production, and procuring a good becomes a matter of searching for an optimal path in the graph from the goods in your possession to the good you really want.

    Finding optimal paths through a sparse graph is a difficult computational problem. It’s much easier if the graph is dense – say, if there’s a small set of goods that can each be traded with almost all other goods, so the path between any two goods has just two steps. You could call the set of intermediate goods that stand in the middle of all the paths “media of exchange” (from the Latin word medium “the middle or center”.)

    That is, in fact, how currencies develop out of barter economies – barterers trying to shorten the number of trades between what they have and what they want. I predict that if your scheme were implemented, the graph would rapidly simplify into a bipartite form with a small group of clearing houses offering to buy and sell anything in exchange for a form of money, and all other participants dealing with each other through a clearing house. Which is how things work now, basically.

    “It is fascinating that you believe that simply introducing a currency magically solves the intrinsic circularity of determining the relative valuations of all goods.”

    Anything that reduces the complexity of economic calculation by as much as the shift from barter to money exchanges does is well worth doing, even if it doesn’t solve the problem completely.

    “Money was the first crude futures contract”

    No. Not remotely. Futures contracts are trades across time – A pays now for B to deliver something later. They are a refinement of loans. A medium of exchange does not, in itself, involve the passing of time in any way.

    “Ian, how is it exactly that you claim to know how people think about how they make purchases? I’m basing it on how I think and some articles I read.”

    Have you tried asking other people? Surely you know some people who aren’t economics nerds. Go ask them whether, when they’re deciding whether to buy something, they mentally translate its price into units other than currency. Ian just told you he doesn’t do so, and I can tell you that I don’t do so – how can you be sure, just from your experience are reading, that you are typical and we aren’t?

    “I see little difference in the scalability, robustness, and simplicity when compared to digital currencies- there’s no contest if you’re comparing digital barter to physical currencies, the former wins in a rout- merely that you are used to the old solution and unused to the new one.”

    You haven’t actually thought through what it would take to implement your scheme – and neither has anyone else, since nothing like it actually exists. When I look at it, with the eye of a software developer, I see it as significantly more complex than currency, and as sacrificing either currency’s scalability or its robustness; it becomes scalable only when unacceptably fragile, and robust only at unacceptably small scales.

    1. >So you envision a massive graph of markets which links each good only to its factors of production, and procuring a good becomes a matter of searching for an optimal path in the graph from the goods in your possession to the good you really want.

      Michael is right. No matter what kind of computational wizardry you apply, such a system would rapidly collapse to some kind of common exchange medium as people tried to optimize out the attention cost of their transactions.

      1. And Nove, before you make the obvious objection; this is true even if a machine is doing the decision making. Because those cycles could be put to better use doing literally anything else, including nothing.

        1. Machines are incapable of decision-making, nor do I see what decisions you’re referring to, so I’d never make such a dumb point. Funny you talk about cycles wasted, considering all the wasted hashing done for bitcoin now. The cost of this scheme is trivial by comparison, though more than online dollars. But just as bitcoin aims to remove the threat of dollar inflation through its distributed setup, this barter scheme does the same more cheaply.

      2. I could try to make some sense out of this if “the attention cost of their transactions” meant anything and if end users had anything to do with such implementation decisions. I have already noted that regional market-makers could standardize on some index unit, but it certainly won’t be currency.

        1. A unit of value accepted for exchange in a wide variety of otherwise unrelated markets … how is that not a currency? It can’t be because it’s not a chunk of some rare metal, or a fancy engraving printed on paper, surely?

  35. With his previous two posts Nove has shown that he does not understand the most basic language or concepts that are relevant here. As such I can now safely dismiss him as an economic imbecile.

    Just for humor sake:

    I have already noted that regional market-makers could standardize on some index unit, but it certainly won’t be currency.

    Calling a currency a shmerp doesn’t make it a not-currency.

    1. >Calling a currency a shmerp doesn’t make it a not-currency.

      /me puts on his General Semantics hat again.

      More generally, you can’t alter the territory by merely changing a label on a map. If the predictive consequences of “X is a shmerp” are indistinguishable from the consequences of “X is a currency”, all you accomplish by insisting that “X is not a currency” is to look a fool.

      This “Nove” character is a crank. He emulates the forms of reasoning and argument, but he is unsane (a GS term of art for a person with persistently inappropriate semantic reactions who is not necessarily insane in the strict sense).

      It’s not that he’s stupid, exactly. He’s more like a religious faith-holder – seems to have an emotionally fixed premise that distorts his thinking near it.

      1. I have explained a couple places above how they may come up with index units that communicate value, but would not be currency in any sense of the word: I can’t help it if you can’t understand it.

        And wow, esr simply calling someone else a crank, the irony runs deep here, considering that is pretty much how everybody else writes you and Stallman off these days.

        I have not fixated on anything: I noted early on that the market may settle eventually on barter matrices or some form of index units. If anything, it’s the silly M. Brazier that fixated on that irrelevant implementation detail. In any case, it appears these ideas are too advanced for this crowd, so no point in engaging further.

        1. And wow, esr simply calling someone else a crank, the irony runs deep here, considering that is pretty much how everybody else writes you and Stallman off these days.

          Take a drink!

    2. Heh, yes, because we all know how finely ingrained “machine… decision making” is in the economic literature. It is hilarious how morons like yourself always make idiotic points and then label others are imbeciles.

        1. @Jeff Read
          “Are you… another Venus Project kook?”

          I had never heard of the Venus Project, so I was curious what that would be. The name “Nove” does sound fitting for the Venus Project. I was put off from the start by this stated aim on their home page:

          It calls for a straightforward redesign of our culture in which the age-old inadequacies of war, poverty, hunger, debt and unnecessary human suffering are viewed not only as avoidable, but as totally unacceptable. Anything less will result in a continuation of the same catalog of problems inherent in today’s world.

          Although these aims are recommendable, war etc. are already seen as unacceptable by the civilized people of the world. This was extensively discussed by Yuval Noah Harari in Homo Deus and Steven Pinker in Our better Angles.

          We do not need to start our culture from scratch for reaching this goal. Worse, “rewrite from scratch” is a sure sign of imminent disaster. I see efforts to rebuild the economy without money as similar misdirected, potentially lethal, efforts.

          Trading status&reputation have always been part of the human condition and are much older than the money economy. But trying to rebuild society on a prehistorical economical systems will produce a prehistoric society with prehistoric demographic numbers (ie, with less 1% of the population surviving).

Leave a Reply to Ian Bruene Cancel reply

Your email address will not be published. Required fields are marked *