By Thomas H. Davenport and Randy Bean

It should come as no surprise that 2018 continued to mark another year in the progression of data adoption in business.  Companies are pushing forward with efforts to become increasingly data-driven.  Firms are investing in transformation initiatives to establish a “data culture” within their organizations.  Early adopters are focused on data-driven business innovation.

As we look ahead to 2019, we reflect on a year of accomplishments and emerging areas of focus – from AI through Ethics (listed alphabetically)

HUD UI Futuristic Frame with Text 2019, Happy New Year. Virtual Background – Illustration VectorGetty

:

  • AI/Machine Learning—AI continued to grow in popularity over the past year, becoming well-institutionalized within many large enterprises. We argued in a previous post, however, that too many companies employed AI pilots and prototypes, and not enough firms had implemented production deployments. As with analytics, the use of AI is increasingly being democratized through automated machine learning (AutoML). Several contributors to KD Nuggets’ review of AI and ML trends for 2019 suggested that AutoML would become more popular over the next year. It will make machine learning models easier to create for business analyst types, as well as dramatically increasing the productivity of data scientists—that is, if they can be persuaded to use it. We also predict that deep learning, which has been the fastest-growing and most popular AI technology over the past several years, will continue to advance in power and prevalence for several years. However, we also expect that deep learning will increasingly be augmented by other approaches to AI. NYU professor Gary Marcus has argued, and we agree, that artificial general intelligence—or even generally useful AI—will have to employ various techniques beyond deep learning in order to be successful.
  • Automation—One trend we noticed in 2018 is that there are a variety of automation technologies for organizations wishing to employ “digital labor” to perform structured work tasks. Robotic process automation, workflow, business rules, process mining, and some forms of AI all have the goal of automating human labor, or at least freeing up humans to do higher-level work. We see increasing numbers of companies embracing these technologies and determining how best to design work to maximize the respective capabilities of humans and machines. Given the proliferation of automation options, it’s important to begin identifying, prioritizing, and categorizing automation use cases so that the right technology is used for each application.
  • Blockchain — 2018 represented a year of major advancement for blockchain solutions as firms sought to ensure that data can be trusted, particularly when managing data in a distributed fashion.  The need to ensure data trust received heightened attention in 2018 due to the adoption of the European General Data Protection Requirement (GDPR), resulting in greater focus on developing trusted frameworks for data sharing.  Healthcare has taken a lead in the adoption of blockchain capabilities, forging ahead with initiatives to ensure the protection of patient data and electronic medical records.  Notably, leading healthcare providers such as Massachusetts General Hospital are evaluating initiatives to store patient data using blockchain. 2019 should be an active front for blockchain in healthcare.
  • Cloud Computing—The cloud continued its march toward domination in 2018. Two Deloitte surveys, for example, indicated that 90% or more of global executives are adopting, considering, or already using the cloud. Amazon Web Services, Microsoft Azure, and Google Cloud are all growing rapidly. They are increasingly adding software and data management capabilities to their clouds, including enterprise data warehouses, advanced analytics, various forms of AI, Internet of Things, blockchain, and robotics applications.  Some companies have converted their computing architecture to 100% cloud. The only question still to be resolved is what large organizations will do with all that well-cooled data center space they are vacating.
  • Cybersecurity—The serious cybersecurity events of 2017—WannaCry and NotPetya—led to many attempts to emulate them in 2018. As data-related activity by good guys grows, data breaches, hacks, and ransomware from bad guys seems to grow even faster. The latest McAfee Lab’s Threats Report suggests that malware exploiting software vulnerabilities grew by 151% in the second quarter of 2018. The volume of these attempts leads us to believe that the only way to address them is through the use of AI/machine learning for cyber-threat intelligence, detection, and resolution. At the moment, however, those technologies don’t seem to reduce the burden on humans much. In fact, they generate way too many false positives. It will be interesting—and important for the world’s data security—to see whether the good guys or the bad guys master AI first.
  • Data Analytics—One of the most prominent and durable trends in analytics is the rise of their use by amateurs or “citizens.” As Tom wrote in a co-authored post for the International Institute for Analytics, graphical and search-based interfaces to analytical programs are increasingly making it possible for those without analytical skills to find data and specify the analytics they need. This opens-up the possibility of data-driven decision-making to many more parts of organizations. This trend started several years ago and will, we believe, continue for many more. If there is data available on a topic within an organization, there will be no excuse for not using it. Analytics in many organizations was also augmented by artificial intelligence, and in many cases a single group supported both technologies. In its simplest forms, machine learning and predictive analytics are basically the same.
  • DataOps – Data Operations (DataOps) is rapidly emerging as a discipline for organizations that continue to struggle with the management of data as a shared business asset.  DataOps brings a set of data engineering principles which borrow from the DevOps software development movement.  The intent is to deliver “rapid, comprehensive, and curated data” to business analysts and decision-makers.  We expect 2019 to be a breakthrough year for DataOps approaches as firms strive to derive value quickly and efficiently from their data assets. We also believe that companies will increasing use machine learning to integrate and improve their data environments, as we described in a post about GlaxoSmithKline.
  • Ethics – Lastly, but by no means to be forgotten, data ethics emerged in 2018 as one of the single most important priorities for leading businesses, stung by security breaches and highly publicized misuses of customer information that represented breaches of public trust.  2018 was in some ways the year that data received a black eye.  Now organizations must rebuild that trust.  2019 can be expected to be a year in which corporations step up efforts to ensure ethical data use and ethical data practices.   As we have noted, the demand for corporate data ethics and greater data responsibility is increasing.  Data ethics is not just good citizenship – it is good business practice. We expect that more companies will add new roles and governance approaches to address this issue over the next year.

[“source=forbes]