With the business world hurried along by disruption, the technology that can quickly deliver valued business outcomes wins the day.
In the field of rising technologies, artificial intelligence is advancing through the technology maturity cycle as companies search for the applications that can quickly improve outcomes, according to Gartner's Hype Cycle for Emerging Technologies 2020 released Tuesday.
Explainable AI in particular shows promise, making its way through the "peak of inflated expectations" and on track to reach the final stage of development in the next five to 10 years, according to the report. Last year, AI as a whole was the backbone in the outline Gartner makes of how emerging technology adoption will change.
"More than a third of the technologies we listed here are related to AI," said Brian Burke, research VP at Gartner. "What I see is that there's tremendous research and development around AI. AI is dominating and becoming pervasive in all the technology that we're using."
AI was a protagonist in last year's Hype Cycle as well, tasked with shaping business and supercharging workforces. But this year, the effect of the pandemic impacted Gartner's future-gazing outlook.
A review of 1,700 technologies, the report identified five key trends that will play out over the coming decade:
- Digital me: Digital passports and social distancing technologies will enable digital representations of people to move past screens and keyboards, using a combination of interaction modalities.
- Composite architectures: The "composable enterprise" can allow companies to respond to a shifting business landscape. Business capabilities are available inside a flexible data fabric, leveraging capabilities such as embedded AI and low-cost single-board computers at the edge.
- Formative AI: A set of emerging AI and related technologies can adapt to respond to changes in the business landscape. The most advanced forms of this technology can produce new models to solve specific problems. Examples include AI-augmented development and generative adversarial networks (GAN).
- Algorithmic trust: Algorithmic trust models replace trust-based models based on responsible authorities. Algorithmic trust helps ensure companies won't be exposed to the risk and costs of losing the trust of their stakeholders.
- Beyond silicon: Technology is reaching the physical limits of silicon, opening the door to new advanced materials to enable the creation of faster and smaller technologies.
The pandemic's effect
Budgets tightening in the context of the pandemic aimed CIOs priorities at areas with the fastest return on investment (ROI). AI is seen as a tool that can help support operations in the face of workforce reduction.
"When we visualize how our business is going to change, how customer behavior is going to change, we'll have to imagine a different way of delivering our products and services," said Burke. "That will require investment in various technologies that maybe there wasn't stomach for before."
If an organization viewed expanding automation within the call center as too risky, or too high of an investment, the rise in call volume faced at the peak of the pandemic left some CIOs with little choice but to adopt automation.
A short-term focus became top of mind for leaders in the context of the crisis. The need to reduce costs and slow down hiring drove executives' actions as they focused on business resiliency and risk mitigation, said Bill Hobbib, SVP of marketing at DataRobot.
"People are looking at 'what can I do in the next couple of months that's going to make an impact,' rather than a long-term strategy," said Hobbib.
The value in explainable AI
With algorithms touching more sensible decision processes — such as whether or not a mortgage is approved or which patient can get access to experimental treatment — there's a heightened need to remove the "black box" of AI.
Increased adoption of explainable AI can aid in compliance. Provisions in the European Union's General Data Protection Regulation give customers a right to ask how AI tools make decisions about them, and ask for human intervention in automated processes.
"Explainability has become a must-have," said Toby Cappello, VP of Data and AI Expert Labs and Learning at IBM.
In assessing the business value of explainable AI, CIOs are making three key considerations, according to Cappello:
- When adopting AI, executives justify ROI with the cost of innovation, viewing AI as "one of those innovative technologies that helps transform operations."
- Introducing bias into processes implies a customer cost, driving them to seek more explainability and insights in their business.
- Explainabiltiy can help deploy AI at scale, ensuring the framework can apply across the enterprise.
But as the technology develops, maturation of AI use will depend on company applications and how siloed the technology remains within the organizational chart, Hobbib said.
