> Linguistic indirection is something of a hallmark of the cultural heritage sector and while it may sometimes be necessary for financial or budgetary reasons it is, in most cases, profoundly harmful or at least a counter-productive distraction and a waste of time.
If linguistic indirection is a term of art, I'm not familiar with it, but it seems like a great way to describe this:
> Digital transformation is the manifestation through commercialization — which is to say financial means and industrial availability — of tools and processes whose introduction shines a light on issues and challenges which were always present but otherwise able to remain unseen.
I may eventually get to the wall label part but this is tough.
> I may eventually get to the wall label part but this is tough.
That wall label was an indirection itself.
To be fair, they did warn at the top: I was asked to speak to you about how [...] AI technologies might inpact the ways in which museum collections are managed. I am going to take a round-about route to get there.
So not wall labels, midway down is: Wall labels, then, are not really the problem. They are the symptom of some broader challenges with the way that museums are organized and the ways in which they get things done.
If you search for those sentences and read the 4 paragraphs above it, you get the condensed version of the problem facing museum data. Basically, they have collections management systems but no one wants to do a bunch of data entry, and when they do, they don't use standards, or consistent naming conventions or semantic labeling for it. And points out: These are not technical problems.
The tie in to how ML/AI can help is a couple paragraphs below it. Basically, please don't use AI to generate narrative wall labels even if the curators are too lazy to organize their collections of researched object information. Also, don't hook commercial LLMs and chatbots to the collections management systems, which contain personal and private donor data. Do use text and image recognition for extracting structured data and object tagging -- for internally use only, and reviewed by humans -- and add it to the museums collection management system.
It was around the "10,000 unlabeled pieces of paper" part that my question became, "is it really important to save all this?" Especially in the context of a design museum that isn't particularly interested in unique works?
I agree that AI should not be used "if the curators are too lazy to organize their collections of researched object information." Just get rid of it. Boom. Done.
I appreciate people that archive and preserve things but that makes a lot more sense when there's like 5 scrolls to be found from an entire century. In the world of infinite data streams there's an almost comical futility to it imo. If the people don't care enough about it, using AI to create more volumes of data on it is just wild.
> If the people don't care enough about it, using AI to create more volumes of data on it is just wild.
They do care about it—they cared about it enough to get it, store it, preserve it. But they’re not good at storing the context around it—it’s like they care about it, but don’t care about why they care about it?
> Curation nowadays is about the purge, the filter.
I agree there’s value in that, but there’s also value in understanding the meaning behind what we keep.
> I may eventually get to the wall label part but this is tough.
Good luck. After the first few paragraphs I thought of a great quote that I heard somewhere: "Twitter ruined my reading skills, but it vastly improved my writing skills."
If you're trying to actually get a point across (vs. writing something that is just read for pleasure) GET TO THE DAMN POINT.
> If we assume that large language models are being used to generate these texts and if those models are able to faithfully and believably parody our long-standing assumptions of what those texts are expected to sound like then does it call in to question the entire practice of intellectualizing an artist's work in their unique voice?
I would say no. Authenticity is always in question. If the artist pasted LLM output wholesale, that was the choice they made to represent their work. Maybe they felt they expressed themselves in the prompt. What if they used a thesaurus, or a ghostwriter, or plagiarized something, or overheard someone say something they liked? It's up to the viewer to decide whether they find it meaningful or resonant.
That's the beauty of art. Intent matters, in that it can affect the interpretation, but ultimately any interpretation is valid.
Pretty interesting. My takeaway is that you can’t assume viewers will know why curators thought a piece was important. The wall label might say what it is and when it’s from but the why (which is the most interesting part) is often missing. And it’s not something LLMs can fix.
Thanks for this summary. Besides adding more text giving the context what can you do? If visitors are genuinely interested then the onus is on them to follow up with books or documentaries.
The point the author was making is that sometimes the information doesn’t even exist in books or documentaries, it’s in assorted documents in the curator’s back office.
What distinguishes an Eames chair on display at the Cooper Hewitt from the same chair on display at MoMA or countless other museums in the world? What distinguishes it from the same chair on display, and for sale, at the Herman Miller showroom?
What, if not the stories that the institutions who collect these objects tell about them?
One of them is near enough to be a visited by me on a day trip. I can understand design museums being essentially franchised showrooms for contemporary culture objects, but I think he asks some reasonable questions about the point of curation and the role of museums in moden society.
Further, to the quote you selected, a glass of wine’s perceived taste improves when you hear the story and see an old or prestigious label on the bottle.
The context does actually change how people experience things. For me visiting a museum is something I do when I’m particularly curious or observant, and the atmosphere typically makes me more so.
”The Cooper Hewitt is a design museum and, like all design museums, it basically has all the same things that every other design museum has.”
Hah, touché.
Cooper Hewitt also happens to be inside Andrew Carnegie’s 19th century mansion on the Upper East Side, E 91st St. It reopens later this week with new exhibitions alongside the amazing house itself, the first floor of which is free entry while installation works are ongoing.
Hearst Castle but with an OG blue-candy iMac in it looking over the Jackie O reservoir instead of the Pacific.
This was a fantastic article. I loved this point near the top:
> Digital transformation is the manifestation through commercialization — which is to say financial means and industrial availability — of tools and processes whose introduction shines a light on issues and challenges which were always present but otherwise able to remain unseen.
And then revisited later:
> Wall labels, then, are not really the problem. They are the symptom of some broader challenges with the way that museums are organized and the ways in which they get things done. I do not think that machine-learning and AI technologies will actually solve any of these problems but the collective hope and belief that they will is, perhaps, the proverbial light I keep talking about making visible some larger issues we have avoided having to address.
I must admit to having had an interest in AI for more than 20 years, and an attraction to museums and their nature for at least 10. I read this article from beginning to end and think it is really interesting. The idea that entertainment gets turned into culture by revisiting? Wow. The idea that museum curators decide what gets revisited (exhibitions with wall labels) and that we just don't keep the data you'd want to train an LLM, and that the very process of training would turn the whole thing beige. There are however lots of uses for ai tools.
My local museum had a room of poorly labeled and unlabeled objects that I loved and so I went down the rabbit hole of identifying and investigating most of them. Then I wrote the missing Wikipedia page for my favorite. If they had provided good labels I would never have gotten so knowledgeable about them.
Adjacently, every plane contains hundreds of social media account owners (ok I'm talking about 737s and larger, not a 4-seater plane). Which makes me wonder if there's a record from every minute of a flight somewhere on social media.
Which in turn makes me wonder if we consolidated all the social media storage, if there'd be a photo (or thousands) timestamped of every moment in time.
Speaking of records of flights, if this was the age of social media overhype (so 2010-2015) someone should've pitched uploading black box data to social media... hah!
Oh wait, it's not black box data, but flightradar24.com, etc, exists...
Super heady.
> Linguistic indirection is something of a hallmark of the cultural heritage sector and while it may sometimes be necessary for financial or budgetary reasons it is, in most cases, profoundly harmful or at least a counter-productive distraction and a waste of time.
If linguistic indirection is a term of art, I'm not familiar with it, but it seems like a great way to describe this:
> Digital transformation is the manifestation through commercialization — which is to say financial means and industrial availability — of tools and processes whose introduction shines a light on issues and challenges which were always present but otherwise able to remain unseen.
I may eventually get to the wall label part but this is tough.
> I may eventually get to the wall label part but this is tough.
That wall label was an indirection itself.
To be fair, they did warn at the top: I was asked to speak to you about how [...] AI technologies might inpact the ways in which museum collections are managed. I am going to take a round-about route to get there.
So not wall labels, midway down is: Wall labels, then, are not really the problem. They are the symptom of some broader challenges with the way that museums are organized and the ways in which they get things done.
If you search for those sentences and read the 4 paragraphs above it, you get the condensed version of the problem facing museum data. Basically, they have collections management systems but no one wants to do a bunch of data entry, and when they do, they don't use standards, or consistent naming conventions or semantic labeling for it. And points out: These are not technical problems.
The tie in to how ML/AI can help is a couple paragraphs below it. Basically, please don't use AI to generate narrative wall labels even if the curators are too lazy to organize their collections of researched object information. Also, don't hook commercial LLMs and chatbots to the collections management systems, which contain personal and private donor data. Do use text and image recognition for extracting structured data and object tagging -- for internally use only, and reviewed by humans -- and add it to the museums collection management system.
It was around the "10,000 unlabeled pieces of paper" part that my question became, "is it really important to save all this?" Especially in the context of a design museum that isn't particularly interested in unique works?
I agree that AI should not be used "if the curators are too lazy to organize their collections of researched object information." Just get rid of it. Boom. Done.
I appreciate people that archive and preserve things but that makes a lot more sense when there's like 5 scrolls to be found from an entire century. In the world of infinite data streams there's an almost comical futility to it imo. If the people don't care enough about it, using AI to create more volumes of data on it is just wild.
Curation nowadays is about the purge, the filter.
> If the people don't care enough about it, using AI to create more volumes of data on it is just wild.
They do care about it—they cared about it enough to get it, store it, preserve it. But they’re not good at storing the context around it—it’s like they care about it, but don’t care about why they care about it?
> Curation nowadays is about the purge, the filter.
I agree there’s value in that, but there’s also value in understanding the meaning behind what we keep.
> I may eventually get to the wall label part but this is tough.
Good luck. After the first few paragraphs I thought of a great quote that I heard somewhere: "Twitter ruined my reading skills, but it vastly improved my writing skills."
If you're trying to actually get a point across (vs. writing something that is just read for pleasure) GET TO THE DAMN POINT.
I suspect the presentation felt different to the audience who were receiving it at the talk, rather than us reading it at our own pace.
The author does note that the article "is, with the benefit of hindsight, the more polished version of what I was trying to say."
But it does feel plucked out of a context/world/tradition that is not as common around these parts.
> If we assume that large language models are being used to generate these texts and if those models are able to faithfully and believably parody our long-standing assumptions of what those texts are expected to sound like then does it call in to question the entire practice of intellectualizing an artist's work in their unique voice?
I would say no. Authenticity is always in question. If the artist pasted LLM output wholesale, that was the choice they made to represent their work. Maybe they felt they expressed themselves in the prompt. What if they used a thesaurus, or a ghostwriter, or plagiarized something, or overheard someone say something they liked? It's up to the viewer to decide whether they find it meaningful or resonant.
That's the beauty of art. Intent matters, in that it can affect the interpretation, but ultimately any interpretation is valid.
Pretty interesting. My takeaway is that you can’t assume viewers will know why curators thought a piece was important. The wall label might say what it is and when it’s from but the why (which is the most interesting part) is often missing. And it’s not something LLMs can fix.
Thanks for this summary. Besides adding more text giving the context what can you do? If visitors are genuinely interested then the onus is on them to follow up with books or documentaries.
The point the author was making is that sometimes the information doesn’t even exist in books or documentaries, it’s in assorted documents in the curator’s back office.
What distinguishes an Eames chair on display at the Cooper Hewitt from the same chair on display at MoMA or countless other museums in the world? What distinguishes it from the same chair on display, and for sale, at the Herman Miller showroom?
What, if not the stories that the institutions who collect these objects tell about them?
One of them is near enough to be a visited by me on a day trip. I can understand design museums being essentially franchised showrooms for contemporary culture objects, but I think he asks some reasonable questions about the point of curation and the role of museums in moden society.
Further, to the quote you selected, a glass of wine’s perceived taste improves when you hear the story and see an old or prestigious label on the bottle.
The context does actually change how people experience things. For me visiting a museum is something I do when I’m particularly curious or observant, and the atmosphere typically makes me more so.
The point is to provide local areas access to such designs in person. What they write on the wall about it is secondary to one's own opinion.
”The Cooper Hewitt is a design museum and, like all design museums, it basically has all the same things that every other design museum has.”
Hah, touché.
Cooper Hewitt also happens to be inside Andrew Carnegie’s 19th century mansion on the Upper East Side, E 91st St. It reopens later this week with new exhibitions alongside the amazing house itself, the first floor of which is free entry while installation works are ongoing.
Hearst Castle but with an OG blue-candy iMac in it looking over the Jackie O reservoir instead of the Pacific.
This was a fantastic article. I loved this point near the top:
> Digital transformation is the manifestation through commercialization — which is to say financial means and industrial availability — of tools and processes whose introduction shines a light on issues and challenges which were always present but otherwise able to remain unseen.
And then revisited later:
> Wall labels, then, are not really the problem. They are the symptom of some broader challenges with the way that museums are organized and the ways in which they get things done. I do not think that machine-learning and AI technologies will actually solve any of these problems but the collective hope and belief that they will is, perhaps, the proverbial light I keep talking about making visible some larger issues we have avoided having to address.
I must admit to having had an interest in AI for more than 20 years, and an attraction to museums and their nature for at least 10. I read this article from beginning to end and think it is really interesting. The idea that entertainment gets turned into culture by revisiting? Wow. The idea that museum curators decide what gets revisited (exhibitions with wall labels) and that we just don't keep the data you'd want to train an LLM, and that the very process of training would turn the whole thing beige. There are however lots of uses for ai tools.
My local museum had a room of poorly labeled and unlabeled objects that I loved and so I went down the rabbit hole of identifying and investigating most of them. Then I wrote the missing Wikipedia page for my favorite. If they had provided good labels I would never have gotten so knowledgeable about them.
What was your favorite?
He kind of lost me with the part about every object having a social media account, every plane at SFO having a social media account.
Adjacently, every plane contains hundreds of social media account owners (ok I'm talking about 737s and larger, not a 4-seater plane). Which makes me wonder if there's a record from every minute of a flight somewhere on social media.
Which in turn makes me wonder if we consolidated all the social media storage, if there'd be a photo (or thousands) timestamped of every moment in time.
Speaking of records of flights, if this was the age of social media overhype (so 2010-2015) someone should've pitched uploading black box data to social media... hah!
Oh wait, it's not black box data, but flightradar24.com, etc, exists...
I started reading, then shifted to skimming, then abandoned it because it has said NOTHING in the first few screenfuls. is this a joke?
[dead]