On artificial intelligence, museums, and hot dogs
--
Inexplicably, it was the “Not Hotdog” app from last season’s Silicon Valley that got — and held — my attention on the topic of AI. The storyline involved an app with one simple purpose: to identify whether an image contained a hot dog or not. If the image contained a hot dog, the app reported “hotdog”. If the image contained no hot dog, the app responded “not hotdog”. For some reason, this app idea made me laugh (a lot) and it had the side effect of renewing my interest in a topic that had previously given me hype and headline fatigue.
It wasn’t long before I started noticing more and more AI-driven experiences bubble up across #musetech conversations. Who didn’t love Google’s Arts & Culture app “lookalike” selfie feature, powered by Google’s hefty machine vision engine? And SFMOMA’s “Send Me SFMOMA” chatbot is a great demo of how AI can open new ways of engaging with collections. (Lots of other examples nicely captured here and here— and the list keeps growing.) Thumbs up to anything that cultivates delight and serendipity in collections discovery!
But something kept nagging at me about AI and how we’re talking about it — and using it — in the museum space. Chatbots, AI-driven machine vision, and voice assistants will evolve their potential as channels for visitor engagement, no doubt.
Yet I couldn’t help but think of parlor tricks, relative to what AI is capable of delivering. What about all the other corners of the museum? Maybe even the organizational corners that drive the collections and outreach and exhibitions and education and ultimately the mission of museums…is there anything there that AI can help illuminate?
What about AI as a set of tools to understand future museum audiences and communities? What else can AI help us predict, classify, and grow?
Museums everywhere are facing the operational challenges of becoming increasingly nimble and responsive to the communities we serve. This takes time. This takes information. This takes sustained attention. This is hard.
Where might AI help us in these operational challenges? Few organizations (and fewer museums) have the resources to fully leverage available audience data to drive strategic, sustained decision-making. But it’s exactly this intersection of audiences, operations, and the museum visit where I’ve been thinking about AI and where it might be used.
Over the next little while, I’ll be pondering aloud in this space on ideas like:
- Applications for AI from the granular (content production) to the massive (understanding communities and demographic shifts)
- Lowering the barriers of entry
- The potential for AI-enhanced predictions and applications for museums
- Next-generation problem sets for AI and machine vision
- …and more
In the meantime, I’ll be futilely trying to resist a real-world hot dog across the street at Shake Shack (I’m partial to the Shack-cago Dog).
Thoughts? Feedback? I’m at @CuriousThirst.
Think piece #2 in the series: On artificial intelligence, museums, and garbage
Think piece #3 in the series: On artificial intelligence, museums, and spaghetti
Think piece #4 in the series: On artificial intelligence, museums, and feelings
Think piece #5 in the series: On artificial intelligence, museums, and authenticity
