The consortium behind New York City’s LinkNYC kiosks is ‘delinquent’ and owes the city millions

I actually came across this back in March and it’s a nice bookend to my post on LinkNYC in general. Key quotes:

Only about 1,800 of more than 4,500 kiosks that were promised by mid-2019 have been installed, Politico reports, with the majority of the kiosks located in Manhattan. Installation has been an expensive, bumpy process that’s fallen short of expectations. Local internet access on the kiosks was revoked in 2016 after officials discovered people were using them to watch pornography in public. Privacy advocates expressed concerns about the way the kiosks could be used to collect data on users. And in 2018, the city agreed to give CityBridge more time to build out the network and share the collected ad revenue.

The consortium behind the LinkNYC program, originally intended to create a free Wi-Fi network across the five boroughs of New York City, owes the city tens of millions of dollars and stopped installing its kiosks in 2018, Politico reports.

Youch.

More on the Limited Capacity of the Internet

I wrote about this at length in The Internet of Pipes, but worth pointing out a few more instances of bandwidth troubles on the internet as a result of COVID-19:

• 88 out of top 200 US cities have seen internet speeds decline this past week, 3 cities by more than 40%
• ‘Azure appears to be full’: UK punters complain of capacity issues on Microsoft’s cloud
• Yes, Twitch is down for some people right now
• Steam is changing how game updates work to manage bandwidth during record usage
• The Internet Is Breaking a Little From the Coronavirus

Awful and terrifying.

“My primary care called from a blocked number, I missed it and they messaged they will call back again. I received no calls. I have chest pain, fever and couldn’t leave the bed the entire day. My symptoms are getting worse.”

Link

This.

Algorithmically generated glitch gloves.

Machine Vision

I’m very interested in machine vision work and where things go wrong there because it seems that everyone is using bad and sloppy data. And when you’re using machine vision to power things like driverless cars, that’s a pretty scary thing! Here’s a few on the subject:

• Some shirts hide you from cameras—but will anyone wear them? (This reminds me so much of William Gibson’s “World’s Ugliest TShirt” from the end of the Blue Ant trilogy.)
• Kate Rose - Adversarial Fashion Sartorial Hacking - Video from DEF CON 27 Crypto and Privacy Village
• How we improved computer vision metrics by more than 5% only by cleaning labelling errors: link and discussion