I once saw a print exhibition where one of the exhibits was a woodcut panel (the kind from from which limited edition prints are made), with a gash cut across it, to guarantee the limitedness of the edition. Is that a norm? If so, is the print or the panel more like an NFT?

I guess the woodcut and the print are like a public/private key pair, except non-cryptographically secured, and limited use. In theory you could, at great cost, invert the print into a very precise new counterfeit block, to generate new prints.

In this case I doubt you could fake it with machine learning. Not only is the training sample too small (say 100 prints in original run) but any production process short of a counterfeit wood block would be detectable as fake.

That’s kinda what I’m trying to get at...like with woodcuts the tech stack to guarantee limitedness is just wood and chisels and a single person could run the guarantee infrastructure using commodity lumber. https://twitter.com/letkma/status/1386360924697595907

But with NFTs there seems to be a deep, dispersed stack of players and technologies which even in the best case is very precarious. Even assuming best case technical execution of contracts, how/where digital assets are stored/referenced, etc.

About half the stack seems solid, the other half, the part encoding how the contracts point to the artifacts and where they’re stored, not so much... see this thread. https://twitter.com/jonty/status/1372163423446917122

Several people have mentioned this — it’s unclear why url pointers are used at all? Why not just hash the artwork in question itself directly on the chain? Content-centric networking seems ideal here. Ditch the addressing problems altogether. Maybe I’m not getting something.

Seems to be primarily a problem for obscure works that, in the stronger case, end up losing the only IPFS host hosting it. Popular works that many hosts have an incentive to keep a copy of seem to be an easier problem. No pointers needed. Just search via hash.

Merely pointing to a URL hosted by a random private party seems clearly like a scam to me. Even if the URL were library of Congress, you’re down to trusted third party as security hole. The whole point of blockchaining things is moot.

Yep, I have seen no good answers, even aspirational/conceptual to this so far. Weirdly, for once, techies seem more skeptical than the artist crowd which is usually at the forefront of the backlash. https://twitter.com/doriantaylor/status/1386367100353060867

I think there’s a there there with NFTs, but it’s technically only a 50% complete invention. And it’s hard to see this because the usual suspect backlashers who tend to spotlight such gaps are now eagerly on board because... for once they’re on the money side of a tech thing

If the premise is similar to my original woodcut analogy, there’s a loooong way to go. Somebody has to invent digital wood basically and make it content-centric discovery. A substrate with a unique digital grain or something (eg . a jpg compressor with a key phrase coded in?)

If the idea is NOT like any of the analogies people have used (baseball cards, woodcuts, autographs, name-a-star scams, diploma mills) then there’s a explanatory gap... what’s the right mental model.

With the original crypto conversations, for many applications, the null hypothesis criticism “this should just be a database, blockchain adds nothing” had a lot of merit. Here the equivalent is “just use DRM.” Or basic PGP signing of slightly bespoke individual instances.

Yes, and my point is those buying in are either too rich to care, too stupid to know better, or treating it as an altruistic priming of an unbuilt future pump, or market-maker types. And the artists I think mostly know this. None of these is a sufficient premise for me. https://twitter.com/simondlr/status/1386371965863272451

To be clear, I’m poking around not because I’m against this but because I’d like to jump on board at some point. I’m usually willing to deal with early adoption jankiness and grow with a medium. But this is an weird case where there’s too much wishful thinking in the water.

I’m also waiting for more interesting distribution protocols on the spectrum between open circulation and must-have-key-to-view decentralized paywall. Like say expanding exclusivity/falling cost. Cost = $ X/serial number, and marginal exclusivity = X/serial number seconds?

So if I price a tweet at $1, first person pats $1 and gets exclusivity for 1 second, second gets 0.5s exclusivity for 50c, etc. I’ve seen schemes like this discussed in auction literature. You own place-in-queue memorabilia, an autographed ticket stub basically.

A smooth microstructured progressively accessible distribution kinda like Hollywood movies (theaters —> dvd (?) —> premium streaming —> basic —> free/cheap archival streaming)

Tbf the art-like market dynamics that prevail now are a turn off for me. I’ve never wanted to make or own original art. Where technical medium constraints (sculpture, oil on canvas for eg) make the original form distinct from digital, and hard to mass produce, I’ll go to museums.

I’m temperamentally digital native, where there is literally nothing special about the “original” unless you impose specialness. But clever distribution mechanisms that ride various pragmatic tradeoffs? Yes.

Maybe there can be artistic merit to clever distribution protocols though. A serial time-rights auction where viewers get access in an artistic rather than random (lottery) or utilitarian (highest bidder) order? An essay that’s released in order of Myers-Briggs types?

That would be fun. An essay of mine that’s released ESFJ to INTP order should behave differently than one released in the reverse order. Or astrological signs. Scorpios first.

Distributional artistry is rare. Banksy for eg. But in a world of aggregation theory, reach-o-nomics, and distribution being more important, maybe go from dumb to smart to artistic distribution?

Distribution as performance art is the only way I can think off to make digital art substantively unique, rather than ascriptively unique with bolted on artificial scarcity and vanity markets driving it. You could genuinely reshape how a potentially historic work makes history.

One reason I’ve kept most of my work free or really cheap is that “people who can pay” is almost never an interesting audience. And the higher the price the less interesting the conversation.

A $5 substack is already a significantly less interesting audience conversation loop than free. A $18 print book is enough to break the loop entirely for all but the biggest works. A $1 million painting in a billionaire’s private collection is basically dead.

(subscribers of *my* various newsletters of course, are uniquely interesting, high IQ, exceptionally attractive, superior humans)

(and I do get that since I’m not dependent on my writing for a living, I can afford to be more picky about how/where/when I play, unlike artists entirely dependent on art, with minimum-wage jobs being the backstop)

In a related vein, I wonder what’s the proportion of NFT-issuers who are:

a) working artists who really need the $
b) people with alt income sources like me
c) trusties
d) primarily crypto people moonlighting as creators
e) other — corporate funded stunt marketers etc

A very good litmus test of how your aesthetic sensibilities shape your commercial posture is to ask, as a thought experiment, what sort of restaurant you’d open? Let’s make this an in-line poll, why not.

I’m definitely on the street-food end. Ideally serving up a “locally world-famous” unique signature item under $5. But not super unique. Something like my spin on a well-known thing. Like a dosa cart with a unique filling.

“Locally world famous” is not a joke. It describes many street foods. Locals take out-of-town visitors there, who become new fans. And people who leave retain strong memories and try to get it shipped to them worldwide at great cost. It’s a good metaphor for subcultural capital.

If anyone ever visits my hometown of Jamshedpur btw, I highly recommend this place. A true NFT among chanachurs. I still manage to snag some on occasion when friends travel there (well, my sister’s friends, since she’s retained stronger ties than me). http://fakirachanachur.com/index.html 

The locally world-famous food metaphor btw points to unique distribution as an angle. There’s often a story to the story of food getting out if local neighborhoods. For eg. friends hand-carrying it on flights, then dividing and mailing the stash as part of care packages.

Once I found a uniquely good potato chip in a DC cafe. Billy Goat chips from St. Louis. Later on a cross country road trip, I specifically detoured through St. Louis for the chips. I’ve since twice special ordered them at shipping cost ~ chip cost. https://buythegoat.com 

Now that’s an NFT among chips. Once when I randomly found a few bags at a Home Goods in the Seattle area, I was elated like I’d discovered a Picasso at a garage sale.

They really are very good. If you’re a potato chip connoisseur, seek them out.

Wonder if this could work digitally. A blog post that’s only findable in a network neighborhood (set of whitelisted requesting static IPs?) but also randomly posted on sites elsewhere. Can you in fact impose such a terroir on a network topology?

Maybe the real innovation needed here is a digital network embedded in a continuous metric space with digital geofencing and exponential drop-off in trust as you retreat from a node. That would be a canvas for NFTs par excellence.

How hard would it be to make a computer network with embedding geography as an explicit topological feature? Like gps on every host and a cryptographically signed location tag on every packet? Otherwise exactly like tcp/ip and interoperable with it as a subnet.

...but allowing for location-based filtering/routing capabilities for qualified packets. The gps need not go with the host but could go with owner.

And geo-mapped hosts need not exist on a mirror world earth. You could map it to virtual worlds.

Ie a local circle drawn on “shitpost tweets planet” should contain people who are geographically close too, but otherwise locations could mean anything, within mapping limits.

This might actually be the best way to do digital real estate. Not like Linden Labs approach. Topological logic based on “proof of proximity” derived from physical earth. Artworks anchor uniqueness via location rather than directly. There may be a million jpegs but with (x, y)

If such a network started and somehow mapped earth coordinates to solar system locations with some freedom, with progressive lock-in, early adopters from LA could choose to set up base on Ceres the asteroid and drag LA there.

How could the math work?

This is an interesting topology problem with limited tearing allowed (the way world map projections like Mercator work... basically an orange peel problem). More freedom than strict projection though. You don’t need shape or distance preserving almost everywhere.

There may be an actual patentable technical idea here. If so, I hereby public-domain it

Let’s see if I can lay out highlights.

1. Every unique content item must be content-addressable and associated with a physical geolocation signed with a private key by the author or rights holder. This tag expires and must be reset periodically or the content piece becomes invalid on the “terroir network”

2. The rights holder may use the location to define access rights for other locations.

3. Every location is mapped 1:1 to a topologically equivalent location on a virtual world, which could be a mirror earth or something else topologically compatible.

4. Every new item added to the world must conform to the topology created by previous items, and will constrain the topological freedom of future items, creating a progressively locked-in virtual geography that’s approximately isomorphic to earth.

5. The progressive lock-in must preserve distance order relationships almost everywhere except along pre-defined global “tear” boundaries, but need not preserve actual distances or angles or areas.

6. The world itself must be uniquely identifiable by its tears/singularities.

This is trivial for any single planet or asteroid. You could also do a set of planets (glue them n-pole-to-s-pole like a string of beads, in order of mean distance from sun, and each join would be a singular latitude on the earth map where the distance metric is discontinuous)

This is fun. I’d build it if I had the technical chops for it. Seems like it should just be a minor extension to tcp/ip stack.

This is no longer an NFT thread. It’s a virtual worlds thread with room for NFT-like inclusions as a natural consequence of proof-of-proximity geographies.

Continuous, orientable geographies are more interesting and powerful than graphs. Few understand this.

See: Blum and Kozen, “On the power of the compass (or, why mazes are easier to search than graphs)” (1978)

One of my favorite papers. https://ieeexplore.ieee.org/document/4567972