Predictions 2020: AI -- It’s time to turn artificial into reality (checks)!

In 2019 53% decision makers have implemented, are in the process of implementing, or are expanding or upgrading implementation of AI.

DQI Bureau
New Update
artificial intelligence

AI is real and ready

In 2019, 53% of global data and analytics decision makers say they have implemented, are in the process of implementing, or are expanding or upgrading their implementation of some form of artificial intelligence.


Twenty-nine percent of global developers (manager level or higher) have worked on AI/ machine learning (ML) software in the past year. Fifty-four percent of global mobility decision makers whose firms are implementing edge computing tell us that the flexibility to handle present and future AI demands is one of the biggest benefits they anticipate with edge computing.

16% of global B2C marketing decision makers plan to increase spend on data and analytics technologies, including AI, by 10% or more this year. It’s clear that many groups across the enterprise have tiptoed into AI. But, to take full advantage , they must overcome challenges in how to prioritize use cases, acquire the right talent, design a governance framework, and choose relevant technologies.

While external market events in 2020 may tempt companies to play it safe with AI, the courageous ones will continue to invest. In 2020, Forrester predicts that:


Twenty-five percent of Fortune 500 will ramp up hundreds of IPA use cases. Intelligent process automation (IPA) combines robotic process automation’s (RPA’s) task automation capabilities with pragmatic AI building blocks, such as text analytics and machine learning.

RPA needs intelligence and AI needs automation to scale. This combination will channel AI to automation-focused outcomes like processing incoming email and documents or supporting employee-facing chatbots in Hr and IT.

These types of IPA use cases, with crystal-clear efficiency gains, will be more appealing to enterprises preparing for the next economic “hunker-down” than those playing the long game investing in transformative AI projects.


In 2020, a quarter of Fortune 500 enterprises will redirect AI investments to more mundane shorter-term or tactical IPA projects; in response, around half of the AI platform providers, global systems integrators, and managed service providers will emphasize IPA in their portfolios.

Three high-profile PR disasters will rattle reputations, but won’t wreck trust in AI. AI can perpetuate harmful discrimination, bias, and even lead to fatal consequences — harming customers and corporate reputations. Facebook, Google, Amazon, Pinterest, and others have already been in PR hot water because of discriminatory or biased AI.

Worse still, if autonomous vehicles struggle to detect pedestrians with dark skin, they can hit and kill one of these undetected pedestrians.


In 2020, the potential areas for harm will multiply: The spread of deep fakes, incorrect use of facial recognition, and over personalization can harm, offend, or creep out customers and employees alike. Fortune 100 firms will have the most to lose, and we expect a few of them to face the most public exposure when things go wrong.

But, these imbroglios won’t slow AI adoption plans next year. Instead, they will highlight the importance of designing, testing, and deploying responsible AI systems — with sound AI governance that considers bias, fairness, transparency, explainability, and accountability.

Confident CDAOs and CIOs will come to the rescue to break data logjams. It’s a common refrain: data scientists spend 70%, or 80% — do I hear 90%? — of their time prepping data before they can even begin to build ML models or gain any AI value. It’s true! data scientists often struggle to acquire, transform, and prepare the data they need to start an ML project.


Data lakes, data engineers, and data prep tools have helped, but the real problem is sourcing data from a complex portfolio of applications and convincing various data gatekeepers to say yes. In 2020, senior executives like chief data and analytics officers (CDAOs) and CIOs who are serious about AI, will come to the rescue, with a top-down mandate to get around the data problem.

Firms with chief data officers (CDOs) are already about 1.5 times more likely to use AI, ML, and/or deep learning for their insights initiatives than those without CDOs. Leadership matters. Next year, more of these confident CDAOs and CIOs will see to it that data science teams have what they need in terms of data so that they can spend 70%, 80%, or 90% of their time actually modeling for AI use cases.

Tech elite will ramp up AI plus design skills, while others will fumble. Today, companies like Adobe and Google pair human-centered design and AI development capabilities. next year, these tech elites will ramp up their efforts to find people with knowledge in both fields — as design skills like problem finding, user research, visualization, and pre-development prototyping prove invaluable.


For instance, Google’s Gmail product teams are honing their skills with techniques to respond to user feedback more quickly and incorporate user data into usability testing, with approaches like Wizard of Oz prototyping. These techniques allow them to learn quickly and surface issues earlier.

But, non-tech companies will turn to designers focused on look and feel and limit their involvement to “design the box around the AI.” As a result, in 2020, just 5% of design or AI job postings will mention the connection between the two fields and many AI efforts will struggle to gain user adoption.

Experienced design providers such as Ideo and Frog have already realized this opportunity to help their clients, and 75% of these providers will formalize an AI center of excellence — though most will focus their efforts on conversational AI.


Four in every five conversational AI interactions will continue to disappoint Turing. Brands have flocked to conversational AI and chatbots to reduce the strain on, and costs of, their customer service organizations. But, overly ambitious projects have failed to resolve customers’ issues or answer their questions much more often than not.

For example, Australian telco, Telstra’s Codi chatbot, generated strong backlash from customers, who resorted to Twitter to vent their frustration. Despite the maturation of the tool sets, including the expansion of prebuilt and vertical-specific intent libraries and higher-power natural language understanding (NLU) engines, by the end of 2020, conversational AI will still power fewer than one in five successful customer service interactions.

That leaves the lion’s share of customers stuck interacting with chatbots that won’t even get close to passing the famed Turing test.

Key takeaways

Gap between AI haves and have nots will widen

2020 will be a year of less experimentation and more implementation. As a result, companies that focus on the right strategy, skills, governance, data, and tools will get ahead of those that don’t think holistically about applying AI across the enterprise.

AI risk-reward tug of war will intensify

Companies want tangible return and reward from AI projects, and yet, will grow wary of going all in for fear of unintended consequences and risks.

-- Srividya Sridharan, Mike Gualtieri, JP Gownder, Craig Le Clair, Ian Jacobs, and Andrew Hogan, with Gene Leganza, Chris Gardner, Diego Lo Giudice, and Chandler Hennig, Forrester.