What TD Leaders Need to Know About AI

As the use of artificial intelligence (AI) grows within learning organizations, many talent development leaders remain unequipped to effectively adopt it. The second session in ATD’s AI discussion series brought together experts to address how TD leaders can incorporate AI into their work. Sae Schatz, founder of The Knowledge Forge; Brandon Carson, vice president of learning, development, and partner experience at Starbucks; and Mike Hruska, president and CEO of Problem Solutions, shared pragmatic examples from organizations that are using AI well, explained the role of data, and addressed attendees’ major challenges. The discussion revolved around these three main themes:

Understanding the Basics of AI

In a poll taken during the discussion, 76 percent of participants identified as AI beginners. The recent rising interest in AI can feel overwhelming, but the software is less intimidating than you might think. In fact, you already use AI in your daily life.

“There is no such thing as ‘AI,’ it's just really good software,” Hruska said. AI is an augmented enterprise designed to take in information, evaluate it, and produce new data based on its prompting. Carson says companies are in the opportunity phase when it comes to AI, and it is important for TD leaders to understand it before jumping into its capabilities.

Using Data to Feed the AI Machine

Essential to using AI is the process of collecting good internal data. “Data is the new oil—it’s raw but unrefined, and it needs to have something done with it,” said Schatz. Handling data is delicate, and AI users should keep accuracy, security, and public trust in mind. Limit risk by only inputting data specific to the AI’s requested task. Be well-versed on the information already in the public domain, and what is considered a company secret. Imagine a sales department, for example. AI can evaluate buyer preferences and suggest products, but it would be risky to allow AI access to a catalog of customer payment information. Leverage data to augment workflow while maintaining public trust.

Claiming a Stake in the AI Discussion

Employee anxiety around AI is high, because it has potential impacts on all types of work. L&D will have a responsibility to upskill and reskill the workforce as tasks get automated and AI becomes commonplace. As a TD leader, address these concerns and invest in people in tandem with investing in AI. “Look forward enough to understand the impact of AI,” said Carson, “and be prepared to keep humanity in the work.”

Find out what policies are in place, and use them to drive your approach to AI implementation. A well-developed policy would include ethics standards, outlines for employee training and use of AI, guidelines for security, monitoring of AI-generated content, and procedures for data collection. If the governance is outdated, be involved in updating new policies. Then, host workshops from the top down to work on literacy, strategy, and implementation.

Barriers, Misconceptions, and Panelists’ Solutions

During the session, the panelists addressed a variety of resounding concerns expressed by attendees about adopting AI.

“My company has banned/limited AI use.”

Company resistance to AI might stem from a lack of understanding of what AI is and how it will be used. Be intentional about what you will use AI for and avoid generic blanket statements. If you can explain AI’s function to others, companies may reconsider their stance.

“AI is a security risk.”

Data protection is a top priority for any company, but AI is not the enemy of data safety. Look into data protection plans that focus on AI specifically. In addition, double check your data before inputting it into a program. Human error can be a bigger risk than machine error. Ask a co-worker to cross-check your work with AI as well.

“AI can’t do a better job than me.”

There is no replacement for human work value, but that does not mean AI can’t do a lot of work more efficiently, thus allowing employees to focus on other priorities. For example, if course design occupies most of your time and limits the number of courses you can offer, using AI as a design aid can reduce that time load so you can invest in your other ideas. “Use AI tools to accelerate work, but do not rely on them as a substitute,” said Schatz, encouraging listeners to switch to an “AI and People” mindset. Use AI to support your work and you will see improvements, but avoid AI out of fear and you may be lost in the shuffle.

“AI can’t be trusted.”

AI requires a user to be well-versed in tech literacy and to evaluate generated content. Be cautious about the content you input into AI and scrutinize the content that comes out, but don’t let this concern stop you from using AI. One tip from Carson is to assign a co-pilot to check your AI-generated content for bias, governance compliance, and quality insurance.

Next Steps Toward Adoption

“You can chase the latest tool for the rest of your life, so start by looking internally,” advised Schatz. Start by seeing which AI tools you have access to within your company. Then, consider the internal problems you face and weave AI in with small, narrowly defined applications using your existing data. Prompt AI to ask you questions, rather than asking it for solutions.

Here are five pragmatic steps to implement AI from Carson and Hruska. If you find using AI difficult, treat the software as you would a young intern and exercise project management and communication skills. If sifting through the noise of AI programs is challenging, research which AI software will fit your needs.

The AI train is accelerating fast, and it is imperative that TD leaders hop on now and get comfortable with AI before it is too late. Hruska’s advice is to start with the workflow pains you experience and deconstruct tasks, not jobs. Prepare for the difficulties of learning new machinery, the cost of implementation, and the learning curve of your team. Then, once you are ready, press the gas pedal.

Read more from Talent Development Leader .