It’s difficult to make predictions, especially about the future, but we can be certain that “AI Washing” will continue to rise and flourish in 2019. That’s what market research firm Forrester calls the hype and hooplah around Artificial Intelligence (AI), the latest set of technologies that is promising to “change the world.”
In its 2019 predictions, Forrester tries to temper the “irrational exuberance for AI adoption” with a dose of reality, looking forward by observing how companies automate their work today while experimenting with adding intelligence—artificial and human—to analyzing data and making decisions. Here’s my summary of two Forrester reports published recently, “Predictions 2019: Artificial Intelligence” and “Predictions 2019: Automation.”
It’s the data, stupid: Most companies will find out that to realize their expectations from AI—exaggerated or not—they must invest in creating “an AI-worthy data environment.” 60% of decision makers at companies adopting AI cite data quality as either challenging or very challenging—it’s their top challenge when trying to deliver AI capabilities.
Automation and intelligence convergence is the new new thing: More than 40% of enterprises will create state-of-the-art digital workers by combining AI with Robotic Process Automation (RPA). The RPA market will reach $1.7 billion in 2019 and $2.9 billion in 2021. By the end of 2019, automation will eliminate 20% of all service desk interactions, due to a successful combination of cognitive systems, RPA, and various chatbot technologies.
No brain, no gain: Two-thirds of AI decision makers struggle with finding and acquiring AI talent, and 83% struggle with retention. The solution may lie, at least partially and paradoxically, in the problem itself: Companies will use AI to find scarce AI talent.
Bringing humans back in: 10% of enterprises implementing AI applications will add knowledge engineering to the mix—human wisdom and expertise—to “extract and encode inferencing rules and build knowledge graphs from their expert employees and customers.”
Humans need not apply but they will reinvent themselves: One-tenth of startups will begin life with more digital workers than human ones. 10% of US jobs will be lost to automation — but the equivalent of 3% of today’s jobs will be created. Automation will contribute to better employee experience as rote tasks come off humans’ plates.
Searching for trust in explainable AI: There will be increasing demand for transparent and easily understandable models. 45% of AI decision makers say trusting the AI system is either challenging or very challenging.
Centralization in, chaos out: 40% of enterprises will have automation centers in place. Change management, unpredictability, control, auditing, and security issues will raise governance concerns throughout 2019. In response, enterprises will invest in central coordination — automation centers — designed on the basis of unifying frameworks.
Another certainty about predictions is that some fail to materialize and other trends suddenly appear from nowhere. I asked the Forrester analysts “What did you expect to happen in 2018 that did not happen and what happened that you did not expect?”
Michele Goetz on AI:
What didn’t happen: I expected companies to recognize that AI was a way to get better business results rather than treat AI as its own analytic project. Instead, firms tended to continue testing and experimenting with machine learning within data science labs. This was true even as most companies were in their second and third year of the AI journey. Thus, AI is still thought of an analytic thing and an IT thing rather than a business accelerator or innovation thing.
What surprised me: Companies collaborated within AI and industry consortiums to grow AI capabilities and competencies over working with their existing service provider partners. So even as they primarily built their AI capabilities on their own with existing solutions or turned to their technology vendors, consortiums is where they felt more AI expertise lay than the analytic service providers and systems integrators that know their businesses.
JP Gownder on automation:
One surprise was the high valuations of robotic process automation companies like UiPath, which raised $225 million in a series C and was valued at $3 billion. My colleague Craig Le Clair has been ahead of the curve in covering RPA and its fast growth trajectory. But the market valuations of UiPath and its competitors are growing even faster than I might have expected.
On the flip side, I expected to see broader adoption of physical robotics by enterprises in both operational and customer-facing scenarios. But there are many barriers: Companies find it hard to coordinate between operations and IT. The engineering problems associated with getting these robots to work in situ are significant. Too many robots still require you to rebuild the workspace into a structured or semi-structured space, while the economics work best for robots that can traverse unstructured spaces. And we lack platforms for easily building systems with physical robotics — a problem I expect to see alleviated in 2019.
For more information, see Forrester’s free guide to its 2019 predictions.
This article was originally published on Forbes.com and was republished on the Attunity blog with permission from the author.
About the Author
Gil Press is the Managing Partner at gPress, a marketing, publishing, research and education consultancy. Prior to gPress, he held senior marketing and research management positions at NORC, DEC and EMC. Most recently, he was Senior Director, Thought Leadership Marketing at EMC, where he launched the Big Data conversation with the “How Much Information?” study (2000 with UC Berkeley) and the Digital Universe study (2007 with IDC). Gil is a regular contributor to Forbes and he blogs on his own sites: What’s the Big Data? And The Story of Information.