Algorithms in the Wild (Sep 9, 2022)
Hope you’re having a good week! Here are a few links featuring real life algorithms from this week…
This company uses computer vision algorithms to automate ike-jime for fishing. Ike-jime is a traditional Japanese method for catching fish in a fast and precise way that better preserves the taste. It requires a lot of skill to do this manually but Shinkei Systems uses machine learning algorithms to do this automatically!
Ike-jime involves piercing the brain with a sharp spike to send the fish to fish heaven, then quickly exsanguinating it, and after that destroying the spinal cord. Gruesome, yes, but all of these things prevent stress, suffering and the spreading of bacteria and destructive substances through the body. But it has to be done precisely and within a couple minutes of the fish being caught, so it doesn’t really scale.
That is, unless you automate it, which is what Shinkei Systems has done. … A computer vision system identifies the species and shape of the fish it is holding, locates the brain and other important parts, and goes through the ike-jime motions, dispatching the fish quickly and reliably. - Devin Coldewey for TechCrunch+
Ford is adding a feature for automatic lane change in its driver-assist system. The update is coming this fall and incorporates algorithms for predicting speed changes and adjusting a car’s position in a lane to avoid other vehicles.
Lane change assist will automatically perform a lane change when requested by the driver tapping the turn signal and can even suggest a lane change if the vehicle appears to be in slow-moving traffic. Predictive speed assist automatically adjusts the speed as drivers approach a sharp curve and will signal the driver ahead of time when a speed change is about to occur. And in-lane repositioning keeps the vehicle in its lane while subtly shifting its position away from adjacent vehicles, especially bigger ones like semi trucks. - Andrew J. Hawkins for The Verge
Interesting read on this man’s vision for using machine learning and natural language processing to create “deep listening” for generating and sharing constructive conversations and stories on a massive scale. Deb Roy points out that a lot of the feedback loops, privacy expectations, and infrastructure from in-person interactions are missing in social media. He has a vision to leverage algorithms to help build the digital scaffolding to bring those things on a massive scale to our digital interactions.
We can design smart tools that provide scaffolding so that anyone can easily learn to facilitate a conversation. Stunning advances in machine learning and natural language processing can be marshaled to create scalable forms of deep listening.
I envision social dialogue networks grounded in human practices of constructive dialogue and sharing of stories, powered by state-of-the art AI and digital design. And I believe the best way to proceed is in collaboration with communities, because it is people themselves who must ultimately have agency in how platforms are designed and used. - Kathryn M. O’Neill for Slice of MIT
“Roughly a third” of Camino Financial’s customers don’t have a FICO score, which can make them invisible to lenders, Salas said. “Recognizing that these are very creditworthy borrowers, we realized very quickly that we needed to develop our own score.” - from AWS Transformation BRANDVOICE
Interesting brief read on the founder of Stitch Fix. Stitch Fix is an online personal-shopping company that uses algorithms to make very personalized clothing recommendations to people. Here’s a digital tour from Stitch Fix on how they use algorithms in their business.
Stitch Fix, which leverages data science and human stylists to send personalized outfits in the mail, has been on a mission to win over customers at a time when a trip to the shopping mall sounds particularly unappealing.
“I founded Stitch Fix to take on a very human problem: How do I find clothes I love?” wrote Lake in a letter to investors at the time of the offering. “Spending a day at the mall, or devoting hours of time to sifting through millions of products online is time consuming, overwhelming and neither effective nor enjoyable. I knew there had to be another way.” - Lauren Debter for Forbes
