Machine learning, deep learning, and other artificial intelligence technologies promise to significantly change the workplace. Reduce redundant tasks, automate work, and improve organizational performance. However, the potential of these technologies is still out of reach for most companies. According to a recent census, less than 7% of organizations have adopted AI technology. why?

An ongoing survey of dozens of companies shows that AI solutions are most often not adopted because leaders are concerned about how AI adoption will affect their companies. increase. They are hesitant because they fear new technologies will displace jobs, disrupt workplace dynamics, or require new skills to learn.

These are not unfounded fears. Implementing AI can bring about disruptive changes and disempower staff and employees. If members are reluctant to adopt new technology, they may be reluctant to use it, resist its deployment, or use it in limited capacity. This impacts the benefits an organization derives from using that technology in the first place. are often unaware that they occur.

But wearing down your nerves and wasting time and resources when deploying new technology is not the answer. Rather, the leader should strategically advance his AI technology deployment. Organizations often spend a lot of resources developing or acquiring game-changing innovations, but don’t think enough about how to successfully deploy them..

To navigate this process, we propose a three-step approach to finding a strategic pace that allows companies to achieve the benefits of AI solutions without enduring the pain of improper deployment. This helps leaders understand who in their organization may be impacted by AI solutions, and uncovers the capabilities and willingness of affected employees to adopt specific solutions.

Step 1: Evaluate the impact of your AI solution

Businesses often struggle to understand who will be affected by AI solutions. They typically conduct extensive research or have managers guess who might be affected, but our research shows that these efforts fall short and overlook an important component. It has been suggested that there are many Instead, we need to get into the weeds to determine which tasks and roles a particular AI solution will change and what the impact will be. Once that is established, it becomes easier to determine what pacing strategy makes sense.

identify the task

To plan how your AI solution will change, you need to understand which tasks it will affect. To do that, we first need to map out how the process is currently done, and then talk with vendors about how it might change. Companies should ask the vendor for detailed specifications of their AI solution and the tasks that may be impacted by their solution. This may require asking tough questions of the vendor. “Where did this technology deployment fail?” “Why did that deployment fail?” , must act as partners.

The point of this exercise is to understand how AI replaces or modifies existing tasks, or introduces new tasks or processes. Identifying new tasks is especially important. Failure to account for these tasks can impact an organization’s ability to deploy AI. For example, many medical systems have tried to automate some of the work done by radiologists only because they realize they need to double-check all AI-generated output.

identify the role

Once you’ve identified the tasks affected by your AI solution, see who’s responsible for those tasks and get a comprehensive view of whose work will change. Missing a seemingly trivial effect can lead to big problems. For example, we observed attempts by hospitals to deploy AI solutions to automate patient scheduling. The company failed to note that it was involved in the patient prioritization process even though scheduling is a small part of a nurse’s role. and as a result nurses opposed a more complete introduction of the technology.

We recommend setting up a matrix that maps impacted tasks by role. To do this efficiently and thoroughly, you need to talk to a manager who oversees the processes affected and who understands who will perform these tasks. Be sure to ask about inputs and handoffs so you don’t miss important intermediaries. For example, when a hospital leader spearheading an AI effort asked her manager in the office about inputs to her scheduling process, she quickly understood that nurses were involved in the process. I guess.

Calculate impact by task and role

Once the matrix is ​​created, assess how deeply your AI solution will impact aspects of your organization. That is, the roles affected and the number of tasks per role. Our observation is that an AI solution becomes very difficult to deploy once it affects at least one-third of the tasks assigned to one particular role. Also If it impacts at least three different roles within the company (even if the solution has only a minor impact on the tasks performed by these three or more roles), in such situations an organization may not consider deploying an AI solution. should consider strategies for pacing

Step 2: Identify barriers to adoption

Not all deployments require a slow pace. However, if your organization is severely impacted, understanding the capabilities and willingness of the affected members to adopt AI is essential to determining the appropriate pace. Leaders should discuss with employees how they feel about the coming changes and consider potential barriers.

Competence barriers are skill-based, centered around whether employees are adequately trained and able to perform the required tasks. Barriers to motivation are more emotional, and in many cases the new technology may ultimately make their role meaningless because it’s either “too good” or “not good enough” to affect performance. It arises out of fear that it might have a negative impact.

The questions may vary slightly depending on your organization’s situation, but we’ve found a few questions that tend to work well. As a general rule, questions are most effective when phrased in a neutral and supportive way.

role-level barriers

If we find that a particular role is going to change significantly, we should look at the individuals who currently fill that role. These questions are designed to help assess the individual’s ability and willingness to adopt new technologies. These conversations are often best held by the team responsible for the deployment.

ability

You need to assess how well the person in that role understands what it takes to use the new technology effectively. It’s okay if they don’t know. I don’t want them to suddenly feel like they’re falling behind or disqualified.

Question: “What should I do new or different with this new technology?”

You’ll also want to learn more about the skills people already have. Ask people to talk about relevant backgrounds and training, and how they feel they are lacking preparation.

Question: “What made you ready to use this technology?”

willingness

Providing a space to talk about how the person in the role feels about the new technology being asked to use can also show that you trust them. Asking simple, straightforward questions sincerely and being interested in their answers is a good way to show respect.

Question: “Do you have any concerns about using this new technology?”

We also need to try to understand how people feel that new technology can affect performance. Are you worried?

Question: “What are the barriers to using this technology?”

task-level barriers

If you have any concerns about how tasks change, it is essential to understand the specific barriers that arise from coordination and interdependence among the people involved in task completion.

ability

A complete system of individuals who must work together to complete the task must be planned. Do the necessary communication channels already exist between these individuals?

Question: “Who do you rely on to complete this task, and how do you work together?”

New AI solutions may require people to adapt to their roles. As a link in a chain, we want to make sure everyone feels they can deliver.

Question: “How will this new technology change the way I interact with others to complete this task?”

willingness

I don’t want to ask people to “out” other employees, but I do want to know if there are any general or high level concerns. Many people are more comfortable expressing their true feelings when speaking generally or about other people.

Question: “Do you think other people are concerned about using this new technology?”

Do you have any concerns or concerns about working with others and new technologies? Try to highlight challenges you may have overlooked.

Question: “How will this new technology affect the performance of this task?”

Step 3: Identify Your Right Pace

Now you know if your employees are capable and willing to adopt AI solutions.rated based on employee responses Ability to recruit When willingness to hire Go to the corresponding matrix quadrant to determine the recommended approach to deployment.

If you are both capable and ambitious, you can fully deploy your AI solution.when Ability to recruit It is low and administrators are advised to use Gradual function slow down the pace. This includes sequencing the rollout of features built into the technology so as not to overwhelm the organization’s ability to absorb new technology.

This has proven effective for insurers looking to deploy AI to help identify fraud. Initially, the compliance team didn’t know how to use AI tools. The AI ​​solution had to be programmed and tuned to take into account regional differences in rules and actions. Managers have deployed only a few use cases to address this issue. These were widely applied and did not require any adjustments. The company paid to upskill his members of the compliance team, who shifted roles from monitoring potential fraud to programming his AI. By introducing technology in more bite-sized stages, managers have enough time to provide appropriate training to individuals in affected roles and to establish the necessary procedures and practices. At the same time, individuals who are asked to adopt new technologies do not have to make sudden or drastic changes, but are empowered to develop additional skills at a pace they are comfortable with.

when willingness to hire low and administrators are advised to use complementary positioning Slow down the pace of deployment and improve employee satisfaction by deploying the solution as a tool for your employees rather than a replacement. This includes easy-to-understand explanations and demonstrations of new technologies so that employees have a clear understanding of how they can help improve performance and make their jobs easier.

For example, AB InBev uses machine learning and AI to help brewers determine when to filter beer. In the future, AI may further automate the brewing process, but today, brewers are using AI as a tool to get the job done. Managers must help employees recognize the future value of new skills and familiarize themselves with new technologies in their career progression trajectories, either within their organizations or in opening up other opportunities in the future. . Where both competence and motivation are low, feature rollouts should be phased and complementary positioning should be used.

When you invest in developing or adopting an AI solution, you need to make sure it delivers the expected benefits. Considering the process and pace of deploying an AI solution is a key aspect of realizing its benefits. Our approach helps deploy AI in a way that achieves benefits over time and reduces the cost of disruption.