Mike Stringer
LightNews LightNews
banner
mike.stringer.io
Mike Stringer
@mike.stringer.io
270 followers 580 following 38 posts
People, data science, and design. Critical friend of technology. Chicagoan.

Currently: dad, fractional leader, nerd
Previously: IDEO partner, Datascope cofounder, physicist, nerd
Posts Replies Media Videos
mike.stringer.io
Same! This prompted me to check www.mediacloud.org, which I think is meant to help with questions like this. It looks like about 1 out of 6 stories about the fires mention climate change (if I'm using it correctly...)
two charts showing the counts of news articles that mention "fires AND los angeles" and "climate change AND los angeles". There is a big jump up on Jan 7, when the fires started. The percentage of news stories mentioning "fires and LA" is about 3%, and the percentage of news stories mentioning "climate change AND LA" is about 0.5%.
January 12, 2025 at 6:25 PM
mike.stringer.io
This paper describes user research on speech recognition software, long before accurate recognition was technologically feasible. dl.acm.org/doi/abs/10.1...

This was written over 40 years ago. Technological feasibility doesn't need to get in the way of figuring out if something can be useful!
Schematic diagram: a typist, located in another room, listens to a participant in a user study and types what they say — simulating a speech recognition program.
January 11, 2025 at 10:58 PM
mike.stringer.io
I jotted down these notes a few weeks ago when I was trying to decide if I want to be on here. Sharing in case they're helpful.

It's OK (but not great) at what I put down as my #1 reason, but feels like it's maybe improving.
Why would I want be on bluesky? Some thoughts: - #1 reason is as a source for links + finding people who make interesting things. I think of it like a high-volume RSS feed of a link-blog that that’s collectively generated by those I choose to follow. I don't feel the need to comprehensively read it: I can dip into it. If I follow someone, it's not necessarily an "endorsement", just that I get an impression that somebody might add something that I find interesting in the moment to that collective, or is part of an interesting community of people. Be intentional about follows, the choice of who to follow is the main way of influencing this part of my information diet. - A sense of the zeitgeist: overall, and in different communities or different topics (vibe-o-meter). - I really like following funny/weird bots, e.g. "Why can't I hold all these?" on mastodon, Youtube glitch. Anything by Kate Compton or Darius Kazemi. FrontPagesBot. - Sharing things I make/write. I made it, might as well share it. I am bad at this and think I'd like to improve. When I do this, it feels good to get a few likes/reposts. Even when I'm writing primarily to myself, putting a link "out there" feels like a sense of closure. - "Tiny acts of goodness”. I share what I personally think is good/valuable/beautiful/true/funny/thoughtful. Some things I consider it a "public service" to amplify: try to include a highlight or some reason why I think it's worth sharing — but really I'm sharing because I'd love for others to read whatever I'm linking to! - I don't engage in all that much dialog, and it hasn't been very common for people to engage with me in dialog about the stuff I share, so this more overtly "social" part of social media isn't very important to me! I have pretty much all notifications off, and don't like to / am not good at being fast with replies (even to text messages... sorry friends). I'm open to changing that, though. - To tinker with the data. I'm interested in information network
December 23, 2024 at 4:54 AM
mike.stringer.io
Vintage (1997) article on generative AI www.nytimes.com/1997/11/11/s...
Screenshot of text: Ultimately EMI's success relies on humans -- not just the composers who wrote the music it imitates but also the programmer-composer who wrote the ingenious computer code. A fair amount of Dr. Cope's own musical sophistication has been woven into the program. And it is Dr. Cope who sifts through the output and picks the compositions that sound good enough to play for the rest of the world. ''I am the ultimate arbiter,'' he said. ''The works I like are released and the ones that I don't are not.'' With Chopin, he said, EMI's hit rate is about 1 in 4, with Beethoven it is more like 1 in 60 or 70. EMI's compositions also benefit from the intuitions of the performers, who know instinctively how to take notes on a staff, whether put there by Mozart, Bach or a computer, and make them sound pleasing to the ear. But for all the help it gets from outsiders, EMI still demonstrates that engaging music can be written by a composer with absolutely nothing to say. Dr. Cope says he does not see why some people find that disturbing. ''When I am camping in the Sierras there is an incredible beauty I see,'' he said. ''But it is unintended by nature. The plants are not trying to express things to me and the mountain is not trying to communicate. But I'm inspired anyway.''
December 11, 2024 at 10:07 PM
mike.stringer.io
An ad from Datamation magazine in 1981. It shows a man sitting at a desk, reacting to a sparkling trail of light emanating from a piece of mail on his desk. His tie is blown backwards. Ack! The text of the the ad says "What the heck is Electronic Mail?"
December 11, 2024 at 8:41 PM
mike.stringer.io
I just updated observablehq.com/@stringerthe... with data from 2024, and with charts showing dry and wet streaks, inspired by @denisedslu.bsky.social in www.bloomberg.com/graphics/202...
A chart showing the number of consecutive days with no rain in Chicago, from 1959 to 2024. #dataviz #datavis
November 12, 2024 at 10:45 PM
mike.stringer.io
feel free to play with these sketchy faces 🤷🏻 observablehq.com/@stringerthe...
three roughly-sketched faces side by side
November 5, 2024 at 8:22 AM
mike.stringer.io
An experiment with something like ridgeline maps, except with dots: here's the Chesapeake bay with (1) randomly sampled dots, (2) dots sampled on a grid, and (3) lines.

Combined day 1 (points) and day 2 (lines) of #30DayMapChallenge
a map of the Chesapeake bay created by sampling points on a grid in an area, and then shifting the position of the dot vertically up proportionally to elevation a map of the Chesapeake bay created by uniform-randomly sampling points in an area, and then shifting the position of the dot vertically up proportionally to elevation a map of the Chesapeake bay created by sampling points on a grid in an area, and then shifting the position of the line vertically up proportionally to elevation
November 2, 2024 at 3:44 PM