Analytics and AI-Driven Business Transformation Is Coming
A recent report from IDC makes some bold predictions about analytics, cognitive computing, and artificial intelligence.
A recent report from market watcher International Data Corp. (IDC) makes some bold predictions when it comes to analytics, cognitive computing, and artificial intelligence (AI).
The first and perhaps boldest of these is that Forbes Global 2000 companies will increasingly monetize some or all of the data they're currently generating as byproducts of their business operations.
For about one-third of Global 2000 firms, revenue growth from sales of this data will be approximately double that of conventional products and services by the end of this year.
"A growing number of organizations will begin to participate in the 'data economy,' where raw data and various forms of value-added content will be bought and sold either in bilateral transactions or through data brokers or marketplaces. Organizations will begin to develop approaches and methods for valuing their data," according to the report, IDC FutureScape: Worldwide Analytics, Cognitive/AI, and Big Data 2017 Predictions. "Although for many organizations the revenue opportunity from data products may not be significant, it will continue to increase in the foreseeable future."
IDC's other projections are surprising less for their content than for their imminence. For example, IDC predicts that by the end of 2018 the vast majority -- 75 percent -- of enterprise and ISV software development will incorporate "cognitive/AI or machine learning functionality in at least one application." Not surprisingly, it predicts this will be true of all business analytics tools.
That's an extremely rapid transformation.
Analytics Moving to the Cloud, Intelligent Assistants to the Desktop
Elsewhere, IDC predicts that by 2018 analytics deployments in the cloud will grow 500 percent faster than on-premises deployments thanks to the availability of new cloud pricing models designed for specific analytics workloads. This will have advantages and disadvantages, according to IDC: "Different pricing models and contractual terms for using specialized cloud analytics services will start to emerge, requiring organizations to expand resources on vendor contract management, usage monitoring, and assessment of appropriate analytics services for specific use cases."
The market watcher also foresees a future in which, by 2019, 75 percent of knowledge workers who work with enterprise applications will be interacting with intelligent personal assistants that "augment their skills and expertise."
Separately, IDC says the use of analytics will eventually permit enterprises to master the problem of unstructured data management. "By 2020," it states, "66 [percent] of enterprises will implement advanced classification solutions to automate access, retention, and disposition of unstructured content, making it more useful for analytics."
A Marketplace of APIs
The IDC report features a few clunkers, however. For example, IDC predicts that by 2019, "APIs will be the primary mechanism to connect data, algorithms, and decision services distributed across digital economy value chains, clouds, and data centers." The API is suddenly sexy, and it's clear that the API as a concept now enjoys a salience that has eluded it for most of its existence.
Yes, organizations are hip to the value of exploiting RESTful function- or process-specific APIs, along with the related challenge of managing different versions of (deprecated or unsupported) APIs in their software development efforts. In the same way, however, APIs are already the primary mechanisms for connecting data, algorithms, and decision services.
APIs have been connecting data, algorithms, and decision services since before the days of the System 360 mainframe. To its credit, IDC's guidance clarifies that it's describing a teeming marketplace of APIs and the challenge of selecting, managing, and versioning said APIs. However, its headline prediction makes little sense without this context.
A Commercial Versus Open Source Reckoning?
Then there's IDC's prediction of an imminent showdown between proprietary and open source software. This projection is based on the assumption that organizations are taking up both free and commercial open source technologies in order to build "agile" alternatives to commercial analytics products that (in too many cases) are encumbered by centralized IT and its adherence to process, policy, and practices.
IDC believes this will lead to a showdown between the "conflicting" needs of developers and nondevelopers. Its diagnosis of conflict is, to some extent, accurate. Its account of the nature of this conflict is insufficient, however.
For example, the IDC report claims that "IT has responded to the need to be more agile ... [by] embrac[ing] open source software ... [which] serves as one vehicle to quickly and cheaply launch a new analytics project and address [the] needs of individual user groups." However, the root of potential conflict has less to do with a preference for open versus closed source software than with the radically divergent worldviews (and priorities) of general-purpose software developers versus data management (DM) practitioners.
In general, programmers prioritize the development of efficient, tested, reusable, and robust code; DM experts, the development of reusable, robust, and scalable data access, integration, and management mechanisms. Generally speaking, many developers prefer open source technologies because they're designed by developers for developers.
The showdown IDC projects has less to do with an existential conflict between the open source and proprietary software models than with the divergent priorities of software developers and DM practitioners. If anything, open source technologies were designed for analytics and programming use cases that commercial offerings were not designed for and have not evolved quickly enough to address.
IDC's report makes a number of additional predictions -- some of them quite provocative. In addition, it identifies several external "drivers" that it says will factor into the decisions organizations will make about cognitive computing, AI, and advanced analytics. The full report is available on IDC's website.