Tesla’s Big Plans, DeepMind Pays For Itself, Internet Drones, and Moore’s Law
This week we review Elon Musk’s big plans for Tesla, how Google uses DeepMind to save millions of dollars, why Zuckerberg is building a fleet of internet drones, and check in on Moore’s Law death watch. Plus, projects to try at home, and our top reads from the past week.
Not a subscriber? Join the Emergent // Future newsletter here.
Musk Reveals Tesla’s Not-So-Secret Plan 🚗
You might have heard: Elon Musk outlined his masterplan for Tesla in blog post. For the past 10-years, Tesla’s vision had been to do:
- Create an expensive, low volume car
- Use that money to develop a less-expensive, medium volume car
- Use that money to create an affordable, high volume car
- Provide solar power
Now, Musk is doubling-down on solar power, Tesla trucks, self-driving cars, and car-sharing — he wants your car to make you money when you aren’t using it. The company has already started developing electric and autonomous trucks and buses.
tl;dr “We’re not an electric car company; we’re a futuristic logistics company and manufacturer.” h/t Fusion
But did you know: Mercedes is testing CityPilot, their semi-autonomous bus program in Amsterdam? The bus recently completed a 12-mile test trip that connected Amsterdam’s Schiphol airport with the nearby town of Haarlem.
Along the route, the bus, which is fully networked and can communicate with traffic lights and other city infrastructure, had to stop at traffic lights, pass through tunnels, and navigate pedestrians.
DeepMind: A.I. That Pays For Itself ⚡️
Google’s latest DeepMind experiment has improved the power usage efficiency in their data centers by 15%, and cut the cost used for cooling by 40%.
Compared to five years ago, Google now gets 3.5 times the computing power out of the same amount of energy.
Essentially, the company Google acquired in 2014 for more $600 million, is now paying for itself. Now that’s electric.
Internet In The Sky 🛰
Facebook just completed the first test flight of their internet drone, Aquila. The drone is part of Zuckerberg’s plan to bring the internet to all 7 billion people on Earth by launching high-altitude, solar-powered drones that beam internet access to the ground.
At cruising altitude, Aquila used 2,000 watts of energy — the equivalent output of five strong cyclists. We wonder: could DeepMind improve upon that?
Facebook HQ says 60% of the global population doesn’t have internet access. And, as many as 1.6 billion of those unconnected live in remote locations with no access to mobile broadband networks.
Moore’s Law is About to Be Reborn 💿
By 2021 Moore’s Law will be dead. That’s the word from the Semiconductor Industry Association, which says that transistors will stop shrinking.
All is not lost, however. Processors could continue to fulfill Moore’s Law by increasing in vertical density.
Meanwhile, Nvidia unveiled their new flagship graphics card: the $1,200 Titan X with 12GB GDDR5X memory.
The company claims it’s 60% faster than previous Titan X. It’s no surprise that Nvidia calls their new processor: “The Ultimate. Period.”
To put Moore’s Law in perspective: 69 years ago, scientists built the first transistor. Today, there’s a graphics card with 12 billion of them.
What We’re Reading 📚
- Chinese scientists to pioneer first human CRISPR trial to treat lung cancer. The team plans to start testing people with lung cancer next month after the clinical trial received ethical approval from the hospital’s review board. (Nature)
- Approaching (Almost) Any Machine Learning Problem. An average data scientist deals with loads of data daily. Some say over 60–70% time is spent in data cleaning, munging and bringing data to a suitable format such that machine learning models can be applied on that data. This post focuses on the second part, i.e., applying machine learning models, including the preprocessing steps. (Kaggle)
- There is no difference between computer art and human art. The algorithmic software was written by a human, after all, using theories thought up by a human, using a computer built by a human, using specs written by a human, using materials gathered by a human, at a company staffed by humans, using tools built by a human, and so on. Computer art is human art — a subset rather than a distinction. It’s safe to release the tension. (Aeon)
- Writing good code: how to reduce the cognitive load of your code. Good code is high-impact, and is perhaps the main reason behind the existence of the proverbial 10x developer. And yet, despite it’s importance, it eludes new developers. (Chrismm.com)
- Code Is Never “Perfect”, Code Is Only Ever “Good Enough”. I want my code to be The Perfect Code, unbreakable, like a flawless diamond in a sea of cubic zirconium. I spend entirely too much time trying to cut that diamond from the rock it’s buried in. But is The Perfect Code even attainable, or should we just settle for “good enough?” That’s the question I’ve been wrestling with this week. (Exception Not Found)
Try This At Home 🛠
- Machine Learning is Fun! Parts 1, 2, 3, 4
- Designing and Implementing a Ranking Algorithm
- A stream processing engine modeled after Yahoo! Pipes
- Connect your house lights to video games with a wifi-enabled RGB LED strip controller
- Build a drawing robot with Arduino and Python
- Turn your deep learning model into a microservice
Emergent Future is a weekly, hand-curated dispatch exploring technology through the lens of artificial intelligence, data science, and the shape of things to come. Subscribe here.
Originally published at blog.algorithmia.com on July 27, 2016.