Joho the Blogmetadata + reality = politics - Joho the Blog

metadata + reality = politics

The US Food and Drug Administration has decided tentatively that meat and milk from cloned animals are the same as from normal animals, so it is not going to require those products to carry special labels.

Too bad.

It’s not that I think cloned food is dangerous. I’m not a scientist, but I believe them, and from what I can see—and, I haven’t looked into this at all, so the following opinion is worth less than the time it’s taking you to read it—cloned food is safe. But that’s not the point. I’d still like the labels to note that the animals were cloned because more metadata is always good. If people don’t want to eat clones for whatever reason, they should be enabled to make that choice. In fact, we’d be better off with full access to the information about what we’re purchasing. Where was the cow raised? What was it fed? What was its weight? What was its body fat ratio? How old was it? Did it get to roam free? Did it have a sweet smile? What was its sign? We’re better off being able to access it all, no matter how farfetched.

But, because of the nature of non-digital reality, taking up label space with a notice that the meat is cloned would itself be metadata indicating that the government thinks such information is worth noting. Metadata in the physical world is a zero sum game.

And that means not only is it true that (as Clay Shirky says) “metadata are worldview,” physical labels are politics. We are forced to make value-driven decisions by the constraints of the physical (labels take up valuable space), the biological (human eyes require fonts to be sized above a certain minimum) and the economic (it is not feasible to attach an almanac of information to every chicken wing). But online, all those limit go away…

…except for the economic. It would be expensive to do a cholesterol count for every slaughtered cow (assuming that cows have cholesterol) simply to gather information that so far nobody cares about, but there’s plenty of information that we’re gathering anyway or for which there is predictable interest—e.g., cloning—that we could make available online (via a unique identifier for each slab of flesh). There would still be politics in the decision about which information to put into the extended set, but it would be a more inclusive, bigger tent, allowing customers to decide according to their own cockamamie values.

And isn’t cockamamie consumerism what democracy is all about? [Tags: ]

9 Responses to “metadata + reality = politics”

  1. Sounds good, but (almost) nobody in the real world wants it. Remember the company which was doing a system of scan a product’s UPC with a cameraphone, use UPC as lookup into database for reviews, send data back to phone? How are they doing?

    Far more important is people who have an allergy to peanuts, finding out what foods contain peanut-derived ingredients (harder than you might think, since certain oils apparently count). As I understand it (I’m not personally affected), they have a very hard time with what’s essentially a database problem – companies don’t supply the data, and it’s difficult to match it to products given the complexity of the manufacturing process.

  2. Seth, I was about to comment about the camera phone concept you mention here, and I think this is a good idea that will take time to gain acceptance, but eventually will.

    And, as the visual recognition capabilities become more sophisticated, a UPC won’t be neede. You’ll just do a snap of the logo on the canola oil bottle, and this will be enough to start a Web search from your phone.

    There will always be those who prefer not to read a label and just choose a product based on the free decoder ring that’s inside the box, but with the ubiquity of increasingly smart cell phones, it can become a habitual behavior among many.

  3. Seth, AURA comes from MS Research. And, yes, you’re right, it’s not doing so well in terms of practical take-up. But LibraryThing (and eventually MSR’s PULP project) let you scan in barcodes of books, and they’re doing ok. This post is talking about a platform, and, as always, the platform is of no value until and unless apps emerge that take advantage of it. I think it’s early to judge the value of the platform.

    But, I actually care more about the post’s overall point than about how much online info we look up about the products we’re about to buy: The economics of the Web takes some of the politics out of metadata.

  4. A lot of fruit is effectively cloned anyway – anything seedless is perforce reproduced by grafting. All the bananas you buy are genetically identical.

  5. David: Actually, I think you’re incorrect – this post is an example of the economics of the Web being PART OF the politics of metadata.

    The problem is that information/attention is limited.

    To claim: “But online, all those limit go away” – No, that’s completely false. In practice, only a tiny, very educated, and well-off part of the population, can *readily* take advantage of all the glitz, and it still takes a significant amount of time.

    But the *assertion* that they go away is a very useful *political* claim.

    In fact, the politics of the metadata is now arguably WORSE, because of the addition of the gee-whiz techno-utopian argument.

    My proof is that there’s a grassroots, sometimes literally life-and-death, example in food allergy, where the users have a strong incentive to create this online metadata system (if it’s wrong, they’re going to suffer and know it immediately). That’s the upper bound as to how well the idea works in practice.

  6. Seth, I don’t understand why your example of people with allergies isn’t an argument in favor of my point. The inclusion of as much data as possible enables groups that have strong needs to create valuable services. For others whose needs aren’t as strong – e.g., people looking for reviews or recipes – services will arise that provide what they need, too.

    I actually have a rather traditional market-based optimism about this, Seth. The masses of data we’ve already set loose on the Web enables a market of info brokers to arise to meet market demands.

    Where am I going wrong?

  7. I’m curious if distributors of non-cloned (is that the best word I can find?) meat and derivative products will be allowed to so advertise:

    “No cloned meat here!”

    “This milk did NOT come from BessieBessieBessie, but from a unique, unreplicated cow!”

    Surely the interested parties (and their lawyers) would want to crack down on such “misleading” metadata…

  8. Here is the core error:

    “that we could make available online (via a unique identifier for each slab of flesh)”

    My argument: the phrase “make available online” [accessed by identifier] is unworkable for a *huge* number of people. The proof is that there are people who have a literal life-and-death interest right now in a type of data, and that is not sufficient in practice to support such a system (tag-lookup-online). Therefore, such a system will not work for lesser concerns such as is-this-meat-cloned?

    There are a tiny, tiny, ultraluxury systems as discussed above. The use of these prototype systems as arguments against this or that label is
    deep politics involving metadata.

    I should note that on the specifics of the particular attribute in question, I think people worried about the attribute “cloned” are very silly indeed (though I grasp their reasoning, I don’t think they’re right at all here).

    “I actually have a rather traditional market-based optimism about this, Seth. The masses of data we’ve already set loose on the Web enables a market of info brokers to arise to meet market demands.

    Where am I going wrong?”

    Well, if I won’t get in trouble for saying this, I’d say the market model used way overestimates the “geekiness” of the overall population, neglects the attention costs and similar “frictional effects”, and passes over implementation difficulties in practice. As well as neglects financing costs for providing the information.

    There’s always a trivial Panglossian market argument – if people want it, it exists, if it doesn’t exist, people must not have wanted it enough. The problem is that’s unfalsifiable, it works for anything. Especially an optimistic version of it: If they want it, it will exist.

  9. I may be missing the point here but ten years ago, when I visited my daughter (who was doing her final year at Univ of Hamburg in Germany), all the vegetables in the grocery stores were labeled by country of origin. I was told this was done to allow the buyer to make a political choice when shopping for food. (We selected lettuce from Israel versus Argentina). German buyers may also have known what chemicals were allowed to be sprayed on the plants by which government; I don’t know. I just remember thinking, why are shoppers in the US not given more information about the food they are paying good money for and consuming? Economists, don’t you argue that the best market is one where there is “perfect knowledge” on the part of the buyer? I’m with Dave – the more information consumers have the better.


Web Joho only

Comments (RSS).  RSS icon