We’re approaching the last year of this past decade and several people are looking back at what futurists predicted 2020 would be like. Of course, there’s still an entire year left before the end of the decade, but if it hasn’t already started to develop by now it won’t by the end of the year. So it’s time to take a look back at the predictions of people like Ray Kurzweil and a few others to see how things have developed and where things took a turn.
To start with, one of the things I always like to do is look back at some of the predictions from years past on what 2020 would look like from a technology perspective. One of my favorites is a concept video by Microsoft about how technology would integrate within our world. In it, you’ll see some things that have come true and some things that just aren’t even close or have been surpassed by a more realistic approach. While the Minority Report boards seem cool, in reality they aren’t the most effective at collaboration. These days, collaboration is really about chat and sharing than anything.
Self Driving Cars
To start with, one big prediction during the decade was that cars would be self-driving. The technology would help alleviate traffic issues and make our streets safer. While the technology hasn’t taken off, development continues and many car manufacturers are ramping up their own research and development groups to try to be the first to make it big time. One company, Embark Trucks, is developing self-driving trucks for the open road, testing these on IH-10, an area with long stretches and moderate traffic.
Amazon is one of the companies testing with Embark Trucks to help move packages from one location to another with little down time. It’s an ambitious goal, but one that could pay big dividends if it’s successful. While there are many self-driving vehicle prototypes today, the problem really is with how the cars perform in crowded areas like cities. Most likely, self-driving vehicles will be restricted to open road areas where driver fatigue can set in and the distances between locations are long.
This particular technology should improve in 2020, but will still not be ready for prime time. However, with each year the technology gets closer to ubiquitous reality. While not delivered by 2020, the chances of this reaching common usage by 2030 are very high. Cities that understand the opportunities self-driving offers instead of building expensive transit solutions will invest in infrastructure designed for autonomous vehicle such as dedicated lanes and improved navigation systems.
Automation Replaces Humans
This is one that continues to improve just as autonomous vehicles have been. The difference is that companies can take advantage of the technology without having to replace all the infrastructure. Robotics can be focused on mundane tasks and with process engineering, more tasks can be assigned to automation, helping improve business productivity.
We have already seen many of the advancements in manufacturing, particularly in the automotive industry. Tour any vehicle assembly plant and you’ll see more robotic arms moving around placing parts on a vehicle than anywhere else. But, even with all that automation you’ll still see workers making sure things are right and doing some tasks that require more precise adjustment or are too complex for a robotic. But, each year the division between human and robotic labor continues to shift towards the robot as advanced learning improves.
The Machine Learns More
One thing you’ll find from the two prior focus areas is the advancements of machine learning and artificial intelligence. This has probably been one of the biggest breakthroughs in the past few years as learning algorithms and chip technology has improved. Each year, researchers seem to find another approach to helping training the programs and improve the quality of output from the machines and programs.
Key to that training is the collection and ingestion of data that is used to train the systems. For example, one of the ways Google used to train its AlphaGo system in preparation to take on Lee Se-Dol in Go was to train the neural network by allowing the computer to play itself. It then combined that tuned network with the search algorithm of Go games to come up with a computer than not only learned from the past, but continued to improve its skills at the game.
ML and AI will probably be the biggest areas of growth going into 2020 and is expected to grow exponentially in the years to come. New chips are being developed that help speed up the neural networks and focus on accelerating training, a task that can often take long periods of time and consume lots of energy. Estimates have shown that training an AI system can emit as much carbon as five cars in a lifetime.
More To Come
There are several other technologies to look at that will help shape our future, but that’s probably best left to another entry. Another aspect I want to look at is how the US is poised in the race to develop and take advantage of these technologies. Often, it seems obvious for our country to delve deeper into a technology only to see that development stifled by political winds. But, for now the future looks extremely bright as the next revolution of industry expands.