Artificial Intelligence (AI) has moved from “when will we do it?” to “how will we speed it up?” in many organizations.
AI passed some important tests during the pandemic, says David Tareen, director of AI and analytics at SAS. “The pandemic put AI and chatbots in place to answer a flood of pandemic-related questions. Computer vision supported social distancing efforts. Machine learning models have become indispensable for modeling the effects of the reopening process.”
“If there’s one reason IT leaders should accelerate the broader adoption of AI, it’s the ability to uncover opportunities.”
But the future upside of AI is still considerable. “Artificial intelligence is designed to reveal what you can’t see due to the sheer volume of data that is available,” says Josh Perkins, field CTO at digital platform company AHEAD. “If there’s one reason IT leaders should accelerate the broader adoption of AI, it’s the ability to uncover opportunities that generate real business value through insights and efficiencies where perhaps there were none.”
That puts pressure on IT teams to deliver and work harder to overcome the challenges that exist in scaling the implementation and adoption of AI in the enterprise.
[ Check out our primer on 10 key artificial intelligence terms for IT and business leaders: Cheat sheet: AI glossary. ]
How to speed up AI adoption and success
We asked AI experts for tips on actions IT leaders can take to accelerate AI use and maturity in their organizations.
1. Begin with the best use cases
“Often, leaders do not know where to begin or bite off more than they can chew,” says Peter A. High, author of Getting to Nimble: How to Transform Your Company into a Digital Leader and president of the technology and business advisory firm Metis Strategy.
“AI and machine learning efforts are best directed at specific use cases, and it may require engaging a broader ecosystem to bring it to life, especially if you have a paucity of AI and ML talent.” Finding great use cases, partnering with business leaders to bring them to life, and engaging with a broader ecosystem for insight, talent, and technology helps, High says.
2. Manage to milestones
“Without clear goals and planned milestones to show progress, AI projects can rapidly turn into discovery.”
“One overlooked challenge with AI initiatives is the time commitment required before tangible results can be delivered,” says Ravi Rajan, head of data science at cyber insurance company Cowbell Cyber. “Without clear goals and planned milestones to show progress, AI projects can rapidly turn into discovery.”
3. Develop not only an AI team but also a playbook
What can you train your team to do internally? Where can you hire new talent that will help on this journey? What external partners will be key to transformation? “Answers to these questions will help develop a more sustainable plan,” High says.
4. Create a multi-pronged approach to skills acquisition
Every business now needs big data specialists, process automation experts, security analysts, human-machine interaction designers, robotics engineers, and machine learning experts. None of them are easy to find. Businesses that want to accelerate AI results need to kick off what Ben Pring and Euan Davis of the research-oriented think tank Cognizant Center for the Future of Work call a “skills renaissance.”
“In addition to having sophisticated hiring and retention plans, organizations need to work harder to leverage the talent they already have,” Pring says. “A root-and-branch reform of upskilling and internal career progression is an important element of the multi-factor HR strategy necessary to succeed at this foundational task.”
5. Invest in data delivery
AI demands good data. It’s crucial to articulate AI-related work in the context of all other activities that need to be in place for an AI project to succeed, says Rajan. That means investing time and resources around data collection, transformation, cleaning, and normalization and managing expectations around the data requirements necessary to achieve AI-enabled business outcomes. “It is absolutely critical,” Rajan says.
[ Want best practices for AI workloads? Get the eBook: Top considerations for building a production-ready AI/ML environment. ]
6. Expand data sources
Ensuring your data is in good shape isn’t enough; businesses also need to bring in richer sets and types of data, says Davis of Cognizant. Start looking at psychographic, geospatial, and real-time data – all of which have the potential to drive better AI-centric performance.
“Managing this data and making it useful for interrogation and leverage by AI systems is an important step on the road to digital maturity,” Davis says. “Without this unglamorous hard work, a lot of data will remain noise and never reveal the signal buried within it.”
7. Consider establishing data tribes
Squads of data stewards, data engineers, and data modelers can swarm around a specific challenge or customer touchpoint.
CIOs and IT leaders who want to accelerate AI are evangelists. (Read also: How to evangelize AI.) Organizations need to spread the mantra of data and AI across every aspect of their operations – not just keep them caged within the IT department, says Pring of Cognizant.
He advises establishing data tribes with squads of data stewards, data engineers, and data modelers swarming around a specific challenge or customer touchpoint. “Executives across functions – not just in IT – should institute a digital culture in which every employee is eager to use and apply these new data services within their roles,” Pring says. Rotating IT staff and non-IT staff between functions helps.
8. Conduct AI performance reviews
“Think of the algorithms that you develop as employees that need to be evaluated, graded, and either promoted (used more broadly, perhaps), demoted (shrinking their application), or fired (taken out of commission if they are viewed as ineffective),” advises High. “Employ a learning loop to continue to refine your practices as you go.”
9. Mind the culture change associated with data democratization
Democratization is the next megatrend for AI as organizations seek to minimize the need for AI subject matter experts, says Tareen of SAS. “Organizations want to reach the next level – cascading the benefits of AI to the masses,” Tareen says. “Customers, business partners, the sales force, assembly line workers, application developers, and IT operations professionals can put AI to work for far-reaching benefits.”
Democratization involves more than access, however. “Often culture tweaks or an entire cultural change must accompany the process,” Tareen says. “Leaders can practice transparency and good communication in their democratization initiatives to address concerns, adjust the pace of change, and result in a successful completion of embedding AI and analytics for everyone.”