New Stanford preprint on AI using images to accurately guess location. Presumably, even if an image contains no geolocation information and it is taken outdoors, a model trained on street view data can figure out where in the world the image was taken.
Nalin is building a new kind of password manager. Definitely going to keep an eye on this project.
I hate so very much how YouTube’s algorithm works. If I watch a single video on a particular topic, my home feed is then completely overrun with content that is related to that thing I just watched. I guess this must work for some people, but it most definitely does not work for me.
I love how Maggie articulates what I believe is the future of the web, or at least a component of what it will be.
Not sure how I missed Spiro World by Time Warp, but discovered it through Yancey’s reflection on 2023 this morning.
I really appreciated Mark’s reflective take on the AI shift.
Inaccuracies in language models are fine when you are creating cool visuals, not so much in spreadsheets that bottom lines are based on. The glitch is the feature (again).
Happy New Year!
From The Internet Is About To Get Weird Again by Anil Dash.
I love the idea of building a world, not an audience. I just discovered Kening Zhu via Rachel Kwon.
