Crystal Ball 2019: ‘AI Transformation Poised to Begin in Earnest’
Years ago, the future was much more opaque. Now, it’s tangible, visible, it’s rising up all around us. A favorite quote from 2018: “I don’t predict the future, I observe it” (futurist Gerd Leonhard). That’s the thing: you can see it taking shape, seemingly in real time.
The future, of course, is AI in all its aspects. AI is at the technology crossroads, the mother of all convergence points. AI innovation is accelerating and AI technology is rapidly gaining in capability. Going slower is AI "operationalization" in the enterprise, but the bulk of high-end IT vendor activity in 2018 was focused on piecing together, falteringly but forwardly, the platforms on which colossal AI power will function.
Also colossal is AI TAM. According to a report from PwC, AI is expected to hit $15.7 trillion by 2030, and a McKinsey report indicated that companies that fully deploy AI over the next five to seven years may double their cash flow by 2030, while companies that are slow to adopt AI may face a 20 percent decrease in cash flows.
So naturally, most of the predictions sent to us for 2019 are AI-related. Let’s take a look at some of the more interesting and, we’d guess, prescient, of them.
The perennial man vs. machine debate brought this observation from MemSQL’s Rick Negrin, VP product management, who subscribes to the belief, prevalent among technology vendors, that AI will put people to work, not out of work. He cited a McKinsey study concluding that new jobs driven by AI investments could contribute about 5 percent to employment by 2030 and that automation of menial tasks will enable employees to focus on more meaningful, valuable work. He also referred to a report in Harvard Business Review that 71 percent of respondents out of 250 executives leveraging cognitive technologies were using AI to automate digital and physical tasks, 57 percent were using AI to identify patterns for business intelligence, and 24 percent used it for engaging employees and customers through machine learning, chat bots and others. (We recently posted an article on the growth of AI in the workplace – based on a survey of workers.)
AI in 2019 will continue to expand from its PoC roots toward enterprise-level AI operations, according to Zachary Jarvinen, head of technology strategy, AI and analytics, OpenText, who cited a Gartner study stating the business value of AI will hit nearly $3.9 trillion in by 2022.
“The long-promised enterprise AI transformation is poised to begin in earnest in 2019,” Jarvinen said. “Most enterprises have reached a point of digital maturity, ensuring access to quality data at scale. With mature data sets, AI providers can offer lower cost, easier to use AI tools for specific business use cases. The effect of enterprise AI at scale will be significant.”
At the processor level, IBM’s Jamie Thomas, GM of strategy and development for IBM Systems, expects a continued move toward purpose-built, accelerated processing to power AI throughput demands.
“The decline of Moore’s Law coincided with the rise of some of the most demanding computing workloads the enterprise data center has ever seen, like machine learning, deep learning, and in-memory analytics,” she said. “These competing realities will drive a transformation of the modern datacenter, where workload-centric architectures replace unconstrained CPU expansion. This will necessitate a focus on architectures and systems designs that maximize IO, take advantage of accelerators like GPUs, FPGAs and high-performance memory.”
AI adoption will require more than advanced solutions from vendors – there’s a burden, according to MemSQL’s Negrin, on CEOs taking a more direct hand in understanding their data infrastructures if AI and machine learning initiatives are to succeed:
…to successfully deploy AI and machine learning to maximize business opportunities and mitigate risks, CEOs and other C-level leaders will need to understand the maturity of their data infrastructures, including how their data is being stored and processed, to determine what technologies and talents are needed for transformation.
Why? Modernizing business means modernizing the whole system…. AI and machine learning…(can) automate processes, identify trends in historical data, provide valuable intelligence that strengthen fast and impactful decision-making abilities. However, the underlying database is what provides the real-time analytics capabilities needed to enable AI and machine learning. It’s pertinent for CEOs and their leadership teams to better understand their infrastructure as that will allow them to prevent wasted investments and time.
This sentiment is echoed in a prediction from Jon Toor, CMO of enterprise object storage vendor Cloudian, whose focus for next year is on data access and management. In 2019, he said, “As businesses increase their use of AI to extract greater value from their digital assets, metadata tagging will become an even more critical element of enterprise storage. This will bring more attention to object storage, which is centered on metadata, and the key will be integrating well with AI tools.”
As machine learning moves from trials to real-world projects, Sivan Metzger, CEO of ParallelM, an ML operationalization
(MLOps) company, foresees an increased focus on decision-making transparency and governance.
“With more ML models going into production every day and an increased focus on compliance, companies will have to account for why models made certain suggestions or predictions,” Metzger said. “In addition, organizations will have to understand exactly where any issues with their machine learning models stem from and why those issues led to certain outcomes. Therefore, we’ll see companies begin to adopt model governance in order to provide detailed fault analysis.”
Akshay Sabhikhi, CEO of CognitiveScale, also looks to the growing role of AI governance – a.k.a. “responsible AI”:
While companies have…increasingly move(d) to operationalize AI within their enterprise, companies are realizing that operationalizing without governance is a recipe for disaster, and can lead to unintended consequences. That’s why 2019 will be the year that organizations wake up to the importance of responsible AI.
At its core, responsible AI is about establishing trust and confidence. In building AI systems that are free from bias, transparent in their operations and able to reflect the core values and policies of the business, companies can implement AI in a more practical, scalable, and responsible manner. If enterprises take a more “people and ethics first” approach to embracing AI frameworks in the coming year, they will be able to unlock the full value of AI.
Doug Hillary, strategic advisor and board member of AI provider Fractal Analytics, sees next-gen networks as a key AI driver: “The launch of 5G in 2018, and the rollout of 5G-enabled edge devices, such as smartphones, hotspots, gateways and IoT devices in 2019, will accelerate the ability to perform analytics in a hybrid or distributed architecture; enabling a new breed of commercial and consumer use cases for AI.”
Jeff Clarke, vice chairman of products and operations for Dell Technologies, agrees, saying “5G will have us livin’ on the edge.”
“It won’t be long before we see micro-hubs lining our streets – mini data centers if you will – that will also give rise to new ‘smart’ opportunities for real-time insights happening on the corner of your street,” he said. “Cities and towns will become more connected than ever, paving the way for smart cities and digital infrastructure that we predict will be thriving in 2030. And it’ll be a game changer for industries like healthcare or manufacturing, where data and information generated out in the field can be quickly processed and analyzed in real time.”
But with AI at the edge, Fractal’s Hillary also sees a dire need to address the AI skills shortage.
“Slowing macroeconomic growth and scarcity of talent, along with an increased focus to reap the benefits of digital transformation in hyper-competitive markets, will force CIO’s to do less in-house, and to rely instead on an ecosystem of partners,” he said. “Executives and HR professionals will need to get serious about addressing the challenge of retraining, or up-skilling, existing workforces, to accommodate, work and thrive with ‘digital labor.’”
AI in 2019 will cut across platforms that impact sales performance, according to Carson Conant, CEO of mobile enablement software company Mediafly.
“We all know that AI and machine learning can identify trends that we’re unable to see as humans,” he said, “but soon, AI and machine learning will identify buying trends across integrated systems, such as machine learning, sales enablement, training and web browsing so that each one informs all. This integration will expand opportunities in both sales and marketing to help provide better experiences and drive results.”
In fact, Daniel Raskin, CMO of GPU-accelerated database vendor Kinetica, sees the coming emergence of a new marketing position: the marketing data scientist.
“For years, marketing has been shifting from a qualitative discipline to a much more quantifiable discipline,” Rasking said. “Marketing data science will become an essential element of every marketing team. The marketing data scientist will derive detailed insight about customer behavior and producing reliable predictive and prescriptive insights based on complex data models and machine learning. These models will evolve from historical analysis into real-time applications that transform how products are delivered to customers.”
Another Kinetica manager, Nima Negahban, CTO and co-founder, sees the rise of another new career discipline, the data engineer, who will be instrumental to bringing AI to the forefront within the enterprise.
“Last year was the year of the data scientist,” said Negahban. “Enterprises focused heavily on hiring and empowering data scientists to create advanced analytics and machine learning models. 2019 is the year of the data engineer. Data engineers will find themselves in high demand – they specialize in translating the work of data scientists into hardened, data-driven software solutions for the business. This involves creating in-depth AI development, testing, devops and auditing processes that enable a company to incorporate AI and data pipelines at scale across the enterprise.”
The service desk – be it, HR, payroll, finance or other department – will be significantly impacted by AI next year.
“Call deflection is the low hanging fruit that can drive significant cost reduction, improved overall employee productivity and experience,” said Pat Calhoun founder of AI service management startup Espressive. “Unlike most technologies out there, the problem being solved is common across most organizations. In fact, 80-85 percent of employee requests are identical at every organization, meaning it is ripe for a pre-built solution. AI driven employee self-help will drive down call volume at the IT help desk, allowing the CIO to reallocate a big portion of his/her budget to strategic projects that are crucial for the organization to remain competitive.”
On the visual AI front, Bret Greenstein, global VP/head of AI at Cognizant Digital Business, said AI cameras will become increasingly capable and real time.
“AI cameras will understand and act on what they see,” he said. “Today humans spend a great deal of time staring at monitors, waiting for something to happen. In the next year, thanks to AI, cameras will be able to understand what they are looking at and not just ‘see’ images. In retail stores AI driven cameras will be spot and prevent theft; on the road they’ll detect drowsy drivers; and, at home, they’ll notify owners of suspicious activities outside their front door… In Cognizant’s research, roughly two-thirds of executives indicate that they are exploring projects that employ computer vision AI technologies.”