IDG Contributor Network: 7 technology trends to watch in 2018
2017 was a memorable year for the tech industry—and not always in a good way. From the massive Equifax data breach to problems at Uber and the challenges with fake news on social media, technology was in the headlines more than I can ever remember.
Thankfully, the world of enterprise tech saw a bit less controversy, but developments were no less significant for businesses. The drumbeat around AI continued throughout the year, and amid all the hype some real use cases started to emerge. Data is now recognized as a primary asset for businesses, which are exploring new ways to connect and analyze data for maximum value. And the move to the cloud continues apace, with enterprises grappling with how to best leverage the diversity of services on offer.
Against that backdrop, here are seven technology trends for enterprises to keep an eye on as we move into 2018.
1. Enough AI for AI’s sake—bring on the apps
This past year we saw a lot of AI for AI’s sake—technologies that, while impressive, can’t necessarily be mapped easily to real business needs. In 2018, the focus needs to shift to building smarter applications instead of just smarter AI. Investments in startups offering horizontal AI technologies—those applied broadly across many use cases—are drying up. A lot of core AI technology is open source and developed in academia. As a result, the differences between one core technology and another—such as natural language processing or computer vision—are not that meaningful. Those looking at AI in 2018 should narrow their focus and consider specific applications that can benefit from it. Ultimately, the goal of AI is not smarter AI but more productive applications.
2. The next winner in the cloud will have a killer data platform
Chances are, in 2018 every organization will have at least some presence in the cloud. But what will differentiate the cloud winners from the losers is going to come down to data analytics capabilities. Data isn’t going to stop being critical to organizations in the next year; it’s going to become more important. But if you look at infrastructure-as-a-service (IaaS) or platform-as-a-service (PaaS) providers, no one has emerged with a superior data environment in the cloud. Even current data-as-a-service (DaaS) tends to be a dumb service, a dataset you can use. The industry needs a solid data platform, and it needs one fast. Some emerging players, like Databricks, have strong offerings, but for heavyweights like Microsoft, Google, and Amazon, winning the cloud wars in 2018 will come down to whoever has the best platform for analytics.
3. We need to put the ‘I’ back in IT
A decade ago, IT people took care of apps, wired up servers, and configured networks, but in the age of the cloud, a lot of that work has gone away. What’s left is data. It’s the key asset for businesses today, and for a long-time management of that data was also the domain of IT. But IT knows little about data science or the new business functions it now supports. If we don’t want the CISO to be the last person standing in IT, we need to put the “I” back in IT—focusing on intelligence and rethinking how IT is structured and the role it plays. Otherwise, the only thing left for IT to do is security.
4. Microservices start to become a liability
The appetite for microservices seems to never end. But despite their obvious benefits, improving agility and helping organizations take advantage of new opportunities, businesses need to be careful in 2018 that they don’t become a slave to microservices. We’ve been through many generations of API building, and today’s loosely coupled model with flexible data representation is an evolution. But versioning is versioning, and the more microservices there are, the more convoluted the system becomes. This is rarely evident in the first generation, and it’s only after you live with it for a while that you understand the beauty of “less is more.” If organizations don’t start to think carefully about the microservices they choose in 2018, they will become their own problem over time, leaving us all looking fondly back at the era of macroservices.
5. Data gravity informs how you build applications
A decade ago, people talked about vendor gravity—there were economies of scale to going with a single ERP vendor, with the promise that everything would work better together. Data gravity is the new vendor gravity. As a result, decisions about collocating applications close to your data replace the appeal of a single-vendor stack. In some cases, data gravity will pull you along—it’s more about the fact that your data is in S3 than it being Amazon.
6. Apache Spark leaves Hadoop behind
Hadoop provided a way to analyze data at scale while ensuring the efficient utilization of hardware, but hardware is not a scarce resource in the cloud. There was always a question of whether Spark needed Hadoop as much as Hadoop needed Spark, and the cloud has largely answered that question. Use of on-prem Hadoop is dead in the water, and the question now is whether Amazon EMR and Azure HDInsight will be the beneficiary—or just Spark without Hadoop.
7. IoT Groundhog Day
It feels like the industry goes around in circles on IoT. While the potential is huge, the most successful IoT applications to date address very specific problems. IoT will continue to advance at a rapid pace, but it’s unlikely we’ll see a ubiquitous platform for widespread IoT application development, across industries and functions, in 2018. Streaming analytics and AI technologies will be the closest thing we have to a horizontal platform. These embedded platform capabilities are critical to identify patterns, optimize behavior, and detect anomalies in IoT deployments without human intervention.
This article is published as part of the IDG Contributor Network. Want to Join?
Source: InfoWorld – Cloud Computing
Recent Comments