Google's Urban Tomorrowland

Google Sidewalk Labs internet city

Project to build tech city from the ground up

Disney opened Tomorrowland in 1955 as an attraction to give park goers a glimpse into the future. But the vision ole Walt had for his Tomorrowland never really panned out. It is replete with a history of cost overruns, half-baked ideas, and disappointing features.

In its current state, the closest thing it gets to displaying a vision of the future is a Star Wars expo. Utopian futures may be hard to realize. But a park showcasing that future may be even harder.

Google and its Alphabet subsidiary, Sidewalks Labs, have plans to make their own version of a Tomorrowland with the hopes of turning it into a livable city. Ventures such as these, imagining and building a utopian vision, don't always end well.

Utopian societies like New Harmony, Freeland, and Fordlandia come to mind. But like most failed utopias, two of these had socialists undertones, so it was easy to see why they failed. And in the other, Henry Ford forbade alcohol - in a mosquito-infested jungle town in the middle of Brazil!

But no such worries for Sidewalk Lab's new city. The plan calls for a total technological city, built almost entirely from the ground up. Almost - because that was the original plan, but of course that was before city planners and politicians had to get involved.

The technological testing ground will take place in the city of Toronto on 800 acres of land slated for redevelopment. Much of the infrastructure is in place, negating the chances of a silicon sanctuary built from scratch. From CNBC  

From the New York Times: They want it to embody the city of the future, a technological test bed for other communities around the world, “the world’s first neighborhood built from the internet up.”

The Sidewalk Labs proposal in the competitive bid for the project floated all kinds of technological dreams: a thermal energy grid that would be carbon neutral, sensors that separate waste from recycling, modular buildings that convert from retail to housing, monitors that track noise and pollution, self-driving transit shuttles, shared-ride taxibots, adaptive traffic lights, delivery robots, heated bike paths and sidewalks that melt snow on their own.

Google's artificial intelligence begetting artificial intelligence 

DeepMind from Google, who brought the world the AI that beat the Go champion, has developed an A.I. software that teaches itself without any human input. It plays millions of Go game over and over again, getting better and better with no help from humans.

Google says it a step toward making general AI systems that can go to work on their own and solve problems in a myriad of industries and human endeavors. One of the main benefits is new AI systems can be built without needing huge swaths of data. 

Simply put, after introducing enough data to get the A.I. processing, the new software system from Google begins to construct models on its own that it then can learn from. From CNBC 

Google AutoML machine learning replicates machine learning

Artificial intelligence systems and machine learning have usually been handcrafted by computer scientists. It is a lot of work to write the code and take time to input data. AutoML by Google is an AI that can replicate machine learning, taking the time-consuming work away from programmers. 

It has proven to build systems that are a little bit better than what humans can achieve. Rather than build neural networks from the ground up, AI machine learning models will already be constructed, fine-tuned by the software. They can be used by programmers who understand them and know how to implement them for all types of applications. From Futurism