I was excited to receive my Narrative Clip this spring. It’s the first consumer lifelogging camera: you clip it to your clothes and it silently takes a photo every 30 seconds. Then you connect it to your computer. It uploads photos to the Narrative server, processes them, and  a while later I can view the photos in my Narrative iPhone app.

“Remember every moment,” the website urges. This device, the website promises, will become your memory. It will capture the moments that are really important to you:

Capture the moment as it happens, without interference. Complement your staged photos of majestic scenery with the intensity of the small moments that matter the most.

narrative-clip-website-image

The assumption that technology can do a better job at capturing human experience than humans can (“without interference”) is a classic example of the dataism I wrote about in my book, and that I talk about in my TEDxBergen talk, “What Can’t We Measure in a Quantified World?” Still, the idea of capturing everyday moments we might not have thought to photograph is intriguing. I was interested to see what my days looked like seen from the perspective of this little camera.

Unfortunately, the Narrative Clip fails utterly at capturing the small moments that matter the most. It doesn’t actually document my life at all. Like all technology, it sees what it sees, not what I see.

Here’s a timelapse video of the 77 photos taken on a Tuesday in Chicago this spring between 15:47 and 17:29. 166 photos were taken in total: these are the photos that the Narrative software thought were the most interesting. The time-lapse video is pretty close to the way you can scan through the still images in the app on your phone.

As the camera was snapping these photos, I was walking with my six-year-old daughter from her school to her ballet class. She was hungry and there were no parks or benches on the way, so we sat down on a low fence that happened to face a wall covered with advertisements. You’ll notice that these are the only faces in the camera, and the algorithms have decided that they must therefore be the most important photographs, using the ads as the cover image for the sequence. Jessie decided she would like to wear the camera during her ballet class, and we thought perhaps the images would be more exciting – all those mirrors and beautiful dancers, you know? But as you see, the camera really didn’t capture anything very memorable about the ballet class either.

It turns out to be really hard to get the Narrative Clip to capture any images of my children, because the clip is worn at chest height and my children only reach up to my waist. So I tried fastening the clip to my jeans pocket instead. Not much better.

According to the website, the software selects the most significant photos using a “momentification” algorithm.

narrative-clip-website-blurb

 

“Momentification” is the process where all your photos are uploaded to the company’s cloud servers, analyzed, sorted and sent back to you with the system’s best guess as to what the most important photos are. Based on my experiences, human faces, even on billboards or on stranger’s faces at the table next to you at a café, are given high priority, which makes sense. Photos that are similar to each other tend to be left out of the time-lapse views of your “moments” so that you get more variety instead of a hundred photos of the same wall.

In my case, the momentification highlighted things that were not important to me, like the advertisement on the wall. Just as importantly, the camera itself did not capture the things that are important to me, like my children. The camera’s fixed position on my chest or jeans pocket gives it a very limited view. Perhaps Google Glass would do a better job, as its camera would move with your head. But even if a camera could perfectly capture what my eyes see, would that really capture my experience satisfactorily?

I wrote more about the Narrative Clip in chapter four of my book Seeing Ourselves Through Technology: How We Use Blogs, Selfies and Wearable Devices to See and Shape Ourselves, comparing it to other forms of life logging. You can buy the book in print or download it for free.

And if you’ve used the Narrative Clip, I’d love to hear how you experienced it.


Discover more from Jill Walker Rettberg

Subscribe to get the latest posts sent to your email.

Leave A Comment

Recommended Posts

Screenshot of a paragraph from a New York Times article published May 12, 2026. Text reads: "The price of tomatoes -tart bursts of flavor in salads and sandwiches — surged nearly 40 percent in April from a year ago on a combination of bad weather, high tariffs and climbing transportation costs."
AI STORIES

Genre glitches and unexpected promotional phrases as a sign of AI writing

A genre glitch is a characteristic of LLM-assisted writing where the text suddenly switches genre, typically inserting a short promotional phrase full of sensory details into an informational text. Genre glitches occur when a word in the generated text is heavily associated with a genre or context that is markedly […]

Top of a ransom note from Shinyhunters hacking group. Text reads: "SHINYHUNTERS rooting your systems since '19 ;) ShinyHunters has breached Instructure (again). Instead of contacting us to resolve it they ignored us and did some "security patches"."
Networked Politics University politics

UiB self-hosts the open source version of Canvas, so wasn’t affected by the breach

On May 1st Canvas announced a security breach, and then yesterday the system was hacked. The login page was replaced by a ransom note: if universities don’t pay up by 12 May, student data will be released. Here’s what the login page looked like yesterday: Way back in 2015, when […]

AI and algorithmic culture Networked Politics

AI-generated images, fascist aesthetics: Dieselbrølet and Heimatstrom

My German is pretty dodgy, so when I first saw Heimatstrom on Bluesky, shared by Roland Meyer, a professor of visual culture at Universität Zürich’s Digital Society Initiative, I misinterpreted it and thought it was a far-right campaign. But no, Heimatstrom is a group of left-wing environmentalists using fascist AI […]

Photo of a billboard ad at Oslo S train station showing a smiliing conductor and the text "Du må ikke sove. Joda, bare sov du."
AI STORIES

“Du må ikke sove”: a floating motif detached from its meaning (or: LLMs can write Norwegian but miss cultural references)

There’s a new ad for the train between Stavanger and Oslo in Norway that uses a line from Arnulf Øverland’s famous anti-fascist poem Du må ikke sove (“You must not sleep”). Du må ikke sove, you must not sleep, the ad says. And then it flips it, jovially, joda, bare […]

Academics in Norway: Sign this petition asking for research-based discussions of how to use AI in universities

I just signed a petition calling for Norwegian universities to use research expertise on AI when deciding how to implement it, rather than having decisions be made mostly administratively. ,  If you are a researcher in Norway, please read it and sign it if you agree – and share with anyone else who might be interested. The petition was written by three researchers at UiT: Maria Danielsen (a philosopher who completed her PhD in 2025 on AI and ethics, including discussions of art and working life), Knut Ørke (Norwegian as a second language), and Holger Pötzsch (a professor of media studies with many years of research on digital media, video games, disruption, and working life, among other topics).  This is not about preventing researchers from exploring AI methods in their research. It is about not uncritically accepting the hype that everyone must use AI everywhere without critical reflection. It is about not introducing Copilot as the default option in word processors, or training PhD candidates to believe they will fall behind if they do not use AI when writing articles, without proper academic discussion. Changes like these should be knowledge-based and discussed academically, not merely decided administratively, because they alter the epistemological foundations of research. Maria wrote to me a couple of months ago because she had read my opinion piece in Aftenposten in which I called for a strong brake on the use of language models in knowledge work. She was part of a committee tasked with developing UiT’s AI strategy and was concerned because there was so much hype and so few members of the committee with actual expertise in AI. I fully support the petition. There are probably some good uses for AI in research, but the uncritical, hype-driven insistence that we must simply adopt it everywhere is highly risky. There are many researchers in Norway with strong expertise in AI, language, ethics, working life, and culture. We must make use of this expertise. This is also partly about respect for research in the humanities, social sciences, psychology, and law. Introducing AI at universities and university colleges is not merely a technical issue, and perhaps not even primarily a technical one. It concerns much more: philosophy of science, methodological reflection, epistemology, writing, publishing, the working environment, and more. […]

screenshot of Grammarly - main text in the middle, names of experts on the left with reccomendations and on the right more info about the expert review feature
AI and algorithmic culture Teaching

Grammarly generated fake expert reviews “by” real scholars

Grammarly is a full on AI plagiarism machine now, generating text, citations (often irrelevant), “humanizing” the text to avoid AI checkers and so on. If you’re an author or scholar, they also have been impersonating and offering “feedback” in your name. Until yesterday, when they discontinued the Expert Review feature due to a class action lawsuit. Here are screenshots of how it worked.