How Will AI Impact Your Talent Strategy?

While much of the discussion around AI in the workplace centers on whether it might allow some jobs to be automated, there are even more cases where it will require roles to be reimagined. That transformation should begin even before AI tools are introduced into an organization, as the demands of effectively implementing these solutions will require new approaches. Organizations should rethink their talent strategy to account for the new skills needed to understand AI use cases, properly deploy AI tools, and drive systemic change throughout the business.

What AI Talent Will You Need?

Keith Collins, CIO of technology company SAS, has said four core types of AI talent are needed to build a successful team to oversee AI projects. They are:

• Someone who understands the business processes that will be involved to achieve the desired real-world results.
• Someone who understands machine learning, statistics, and analytics who can ensure that the right techniques are being considered.
• Someone who understands how business data is produced and why it’s used, as well as the quantity and quality of data available in your organization and from outside sources.
• Someone who understands the AI and machine learning architecture required to create your desired business outcome.

In some organizations, one person might fill more than one role, and in others, a role might be divided among an entire team. The important thing is for the business to be able to trust its expertise in each of these areas.

Conducting an AI-Centric Talent Audit

Once you’ve identified a role for AI in your business and defined the results you want, you’ll need to align your talent strategy with your desired transformation. The first step is to look through your organization to identify employees with the skills to manage the integration, launch, evaluation, adjustment, and ongoing monitoring and maintenance of the system.

Ram Narasimhan, global head of AI and cognitive services at the Atlanta-based consulting firm Xebia, has identified a set of key roles required for an effective AI team. Not all organizations will necessarily need all of these functions, but it’s worth considering each role in the context of an organization’s needs.

• Chief AI or technology officer: This is a top technical officer or other senior executive overseeing the project, responsible for ensuring that the AI solution stays in line with the business’s strategic mission. Whoever is in this role needs to understand the rapidly evolving AI space well enough to guide other leaders through what is possible with current technology. They also need to be able to make effective business cases for investment and serve as a bridge between technical staff and nontechnical leaders.
• Business analyst(s): People in this role translate the business needs, processes, and limitations to be incorporated into the AI solution, initially reviewing the project guidelines to understand how they will impact different areas of the business. People may be added to serve in this role for different phases of an AI project, with the main purpose being to help turn their business challenges into tasks for the AI team. For example, while a call center director won’t know the data attributes within the systems they use, they can explain the problems they face each day.
• Data engineers and scientists: The successful application of AI depends on the quality of data and algorithms within it. People in these roles will define how data integrates into the overall AI process, as well as focus on extracting information from several sources and developing it into algorithms that can help make business decisions. Much of their time might also go into cleaning data, such as detecting and removing duplication, blanks, and anomalies. These roles will typically benefit from programming skills and data science backgrounds.
• IT/systems architects: This role augments the work of data engineers and scientists to integrate AI projects into the organization’s existing technology. Usually, this will be somebody with an IT background focusing on system integration and understanding of the existing technology, with a working understanding of AI models and the corresponding IT requirements.
• Systems developers/DevOps: They work with the data engineers, data scientists, and IT architects to make sure their solutions can be turned into viable technology. Also, they likely have worked with traditional programming languages such as Java and C++.
• User experience/marketing experts: Someone with marketing or digital teams experience who understands the demands of the organization’s consumers and other end users. Marketing experts typically are valuable in making sure the project offers the right products or services, whether to internal or external customers. While the technical team might create a brilliant innovation, it also needs to work for the customers or employees who are expected to use it.
• Domain/subject matter experts: These team members contribute their understanding from each phase of the business areas where the AI project will be deployed. For example, if an algorithm will be used to make a product decision, it needs to be validated by a product manager to ensure it makes practical sense.
• Risk and compliance experts: These team members analyze the AI projects to make sure they won’t breach any industry or government guidelines.

As you identify gaps in your organization, you can begin to see where investments will need to be made. But increasing your headcount isn’t your only option.

Finding AI Expertise: Train, Hire, or Outsource

Once you assess the potential costs and what kind of budget your organization can devote to launching an AI initiative, you can determine what roles can be filled with existing employees, whether to retrain existing employees or bring in new hires with the appropriate tech skills, or whether these functions should be handled by consultants or vendors. For small or nontechnical organizations, the best approach might be purchasing an off-the-shelf solution or bringing in a software-as-a-service (SaaS) vendor to handle the project.

Whether your organization is big enough and well-versed in technology to implement your own AI program or whether you entirely outsource the project, you will also face the prospect of training your existing workers and developing employee buy-in.

The first step is to educate members of the organization about the basics of what AI is and is not, what it can and cannot do, and how it is already being used. Share examples of how AI use in other organizations has improved productivity, reduced brain-numbing busywork, improved in-house functions (such as fielding HR questions or generating training recommendations), and otherwise contributed to the workplace.

But, like any new technology or companywide change, implementing AI will require learning new skills, adapting existing procedures, and simply becoming trained (where necessary) on how to work with it. One point to emphasize to employees is how becoming trained and knowledgeable about AI and aspects of machine learning can add to their skill set, make them more valuable and competitive in the marketplace, and set them up for new paths in career advancement.

Tips for Training Your Team on AI Tools

Stephen Chen, Chief Technology Officer for NuCompass Mobility Services, a global relocation management company, recommends three actions businesses can take to train current employees in AI:

• Educate your employees about the technological changes AI will bring, and build a culture of positivity around the technology.
• Prepare your employees for AI reskilling and upskilling by encouraging a “growth mindset” in which they learn to welcome challenges.
• Identify AI use cases for automation that are worth the effort, and eliminate those that are not.

In a survey conducted for the employee training software firm TalentLMS, 49% of U.S. employees say they’d need training to be able to effectively use the widely distributed ChatGPT program and similar tools. Of the remaining respondents, 23% said they didn’t need training because AI tools were easy to use, and 14% responded that they didn’t plan to use AI tools at all.

Any change, of course, is accompanied by some fear, which must be addressed in training. Stress the ethical and other safeguards being put in place by your AI deployment team; the work of your AI ethics committee (if your organization is creating one); and the relevant AI ethics training the organization is conducting.

The specifics of workplace training will depend on whether the organization is fully developing its own full-blown AI solutions, outsourcing some portion of the project, or handling the entire process with an off-the-shelf AI solution. Sometimes, vendors can supply training on the specifics, outside training programs can be added, or the organization can create its own internal training program. AI considerations and specifics will need HR leaders as the primary drivers of the discussion in consultation with other leaders.

Remember that the process of re-evaluating your talent strategy isn’t a set-it-and-forget-it affair. This is a rapidly evolving space, and you’ll need to continually update your plan as new use cases emerge and new skill sets become essential.