2 Emerging Technologies Your Business Needs to Discover: Machine Learning and Connectivity of Tomorrow
What emerging technologies does your business need to know in 2019? Two of the biggest are machine learning and low-earth orbit connectivity of tomorrow.
Technology is always changing, and so there are always new frontiers of digital innovation that your business needs to understand so you can make smart choices on how to leverage them to solve real-world problems in your company.
In this post, we’re going to discuss some of the emerging technologies that will create digital disruption in companies in 2019.
We’ve written in the past about topics like Internet of Things, blockchain, and quantum computing. IoT solutions have moved beyond emerging technology to be more and more frequently deployed in manufacturing, logistics, and other industry verticals. This technology isn’t emerging—it’s here.
And our Worthwhile team is hearing more and more about blockchain solutions in the market that go far beyond Bitcoin to solutions that provide traceability and provenance solutions. Most blockchain solutions are not yet fully mature, but they will be soon. And it appears that blockchain applications will provide a disruptive change to digital innovation. Deloitte’s 2019 Tech Trends report put it this way: “Today, blockchain is to trust what the web was to communication: a profoundly disruptive technology that transforms not only business but the way humans transact and engage.”
How much of a difference could it make? In 2018, the state of West Virginia tested a blockchain-backed absentee voting system. And in our hometown of Greenville, S.C., a startup called Bandwagon is using blockchain to revolutionize the way event tickets are bought and sold. These examples show that blockchain is continuing to expand into more and more horizons and software applications.
Of course, blockchain isn’t the only emerging technology that will fundamentally change what software can do. So in this post, we are going to examine two other solutions that are starting to find real-world digital applications in businesses like yours:
Connectivity of Tomorrow
You might not have heard of machine learning, but you have almost certainly heard the hype about one of its primary outputs—artificial intelligence. AI is such a popular buzzword that evenpolice departments are buying into snake-oil promises of what it can do.
Artificial intelligence is certainly a sexier term (we should probably thank Haley Joel Osment and Steven Spielberg), but machine learning is the real emerging technology that chugs along behind the scenes to create solutions that can be successfully deployed in the real world here and now.
Machine learning takes the toil out of data analytics and forecasting. It can drive decisioning engines that do things like claim adjudication or application approvals. It can use data collected from sensors on the manufacturing floor and predict when a machine will need maintenance or even predict failure. The power of machine learning is immense.
Back in 2005, professor and thought leader Geoffrey Bowker said something that has become a siren call for data scientists everywhere: “Raw data is both an oxymoron and a bad idea; to the contrary, data should be cooked, with care.” In this example, machine learning helps you go from a sterno flame to a convection oven so you can cook your data more fully and more quickly, and thus run your business better.
Two caveats to consider before you sprint headlong into a machine learning project for your business:
First, machines need time to learn. The early outputs of any machine learning system will not add much value to your business, and they may be wildly inaccurate. But if you trust the practice, the machines will learn as the data lake it analysis grows, and you will reach an exponential point of machine learning accuracy, to the point where you can strongly trust any prediction it makes based on your data. Expect this learning curve so you can reach the benefits.
Second, make sure that you ruthlessly eliminate bias from your machine learning algorithm. Because machine learning evaluates patterns in data, your software will need to be told what inputs to trust and which not to.
One wrong input can cause false positives or negatives, and the machine will learn from these false results and actually become more inaccurate. For example: Amazon thought it could make hiring better through an AI project that used machine learning to analyze candidates based on the successful employees the company already had. Noble idea, but it didn’t work because Amazon’s existing employee base was mostly male. The machine learning system included gender as a criteria, and concluded that males were more successful than females, and therefore taught itself that it should favor male candidates.
Amazon thankfully realized the error in the system before taking it into production, but the bias caused by having the wrong inputs shows the limits of what machine learning can do. Companies need experienced data scientists to limit the machine learning process to only the data that should contribute to results.
This highlights the source of problems when it comes to ethical uses of artificial intelligence and machine learning. Benedict Evans, a partner in Silicon Valley venture capital firm Andressen Horowitz, puts it this way: “The 'AI ethics' problem is not so much anything done at Google—it’s things done by third-tier vendors who don't really understand the science, salespeople who don't care, and unsophisticated buyers who hear 'AI' and imagine this is HAL 9000. Of course, this is exactly the same problem we had with databases, and punch cards: people being people.”
Connectivity of Tomorrow
You’ve probably heard of machine learning by now, because the term has been in the ether for a while. On the other hand, connectivity of tomorrow may be a new idea to you. But it has the opportunity to be transformational just as machine learning will be.
What is connectivity of tomorrow? It starts with 5G cellular connections but also includes low-earth orbit satellites, or LEOs. These satellites are part of what will power the new 5G networks that are rolling out across the world.
This innovation is necessary because of how disruptive the Internet of Things revolution has been. The ability to monitor or control things through connected devices started with smartphones, but before long it will be affecting self-driving cars. It allows food companies to track their products from the field to the truck to the production facility to the warehouse and to the store loading dock. And these are just a few examples.
All of these newly connected devices create exponentially more amounts of data. Professor Theodore Rappaport of the Tandon School of Engineering at New York University described the situation aptly:
“The amount of data consumed globally increases by 50 percent each year, and I predict that four years from now our current 50 percent annual ramp will reach 70 to 80 percent. Why the jump? Because the rollout of 5G will accelerate data consumption exponentially. Organizations continually need wider pipes to accommodate ever-growing data volumes.”
As your business wants to collect more data, and as your business uses more and more sensors to collect it, then you will need better and better connectivity options. And it needs this connectivity of tomorrow through 5G and LEO before you reach capacity on your current network with the amount of data you collect. Now is the time to ensure that, as Professor Rappaport put it, the connectivity pipe your business uses is wide enough to handle the flood of data that’s coming. After all, if you can’t connect to devices to collect data reliability, you won’t even have the chance to reap the benefits of data analytics in your organization.
Conclusion: What’s Next for Your Business?
When it comes to emerging technologies-—whether it’s machine learning and connectivity of tomorrow, or the technologies that will enter the hype cycle in 2020 and beyond--it’s vital that you ask the right questions.
Where, specifically, can we use this technology in our business to solve a problem?
Which approach is more suited to our needs and culture—incremental vs. disruptive innovation?
What ROI do we hope to gain?
With the right approach (and maybe the right partner, like Worthwhile), your company can enjoy the benefits of adopting emerging technology to solve business challenges.