Three Mistakes That Derail AI Adoption

Here are the top three mistakes I see some organizations making while attempting to get AI adopted by their workers (later, we will also see more effective ways of doing so).

1. They just focus on providing a tool without also ensuring it gets used and used well.

2. They do not see supervisors (their buy-in and their leadership skills) as the #1 determinant of whether teams adopt new tools and processes.

3. They focus on using AI on a worker’s core tasks, forgetting about its ancillary ones (which are often better targets).

Let’s examine each point and provide examples and solutions.

1. Examples, examples, examples

We can divide the population into four types based on how eagerly they adopt a new tool or technology, such as AI:

• Alpha users: just give them the tool, and they will figure out how to use it.

• Early users: they need both a tool and relevant and concrete examples of how to use it.

• Late users: in addition to the above, they need to see other people like them to use the tool successfully.

• Resistant users: they won’t use the tool until forced to.

Unless you deliberately configured your hiring process to only target alpha or early users, your organization probably contains all four types.

Your AI adoption strategy should take this into account. Hence, it should not only make AI available to workers but should also give them relevant and concrete examples of how to use it, ideally in person by their direct supervisor or peers.

Note that it's not enough to provide generic training – such as "how to use this AI tool regardless of whether you're a project manager, salesperson, or engineer." People need relevant and concrete examples.

As a rule of thumb, if your training material could be given to multiple roles within your organization, it's not relevant enough and won't drive the adoption of the new tool.ShareCopy

2. Get supervisor buy-in and train them on training others

People won’t change their ways just because their CEO asked them in a corporate-wide email. They need to see that their direct supervisor is actively committed to it, too.

If their supervisor doesn't use AI, they won't use it either.

If their supervisor doesn't care daily about whether they use AI, they won't care either.

Supervisors are the lynchpin of any tool or process adoption strategy.ShareCopy

You must work closely and frequently with them to ensure they are committed and transform this commitment into visible yet genuine actions – such as using AI themselves or coaching individuals into integrating AI into one of their sub-tasks (more on this in the next point).

Moreover, you must ensure they can do that in a way that is helpful and effective – as opposed to cringe or meddling. You cannot rely on them already knowing how to do that – you must train them.

3. Focus on subtasks

There are some jobs where AI can easily and effectively help with the main task – translators, some coders, etc. What about the rest, though?

For most jobs at the moment of writing, the core task cannot be delegated to AI yet – AI is not good enough, or maybe human touch is still a requirement, as in most sales jobs, or perhaps AI would be mostly ready, but we cannot tolerate errors. Either way, when the core task of a job cannot be delegated to AI, it is tempting to default to “therefore, that job should still be 100% performed by people.”

However, there are alternatives. What about the non-core tasks of a job? For example, a salesperson might still want to conduct sales in person but use AI for repetitive tasks, such as administrative ones, or to prepare for a sales meeting.

Or what about sub-tasks? For example, maybe you want to keep a human in the loop of a critical task, but there are parts of it that can benefit from AI.

Either way, my advice would be not to treat jobs as a monolith but rather see them as an ensemble of tasks and consider whether any subpart of it can benefit from AI – and not necessarily for full automation but either partial automation or augmentation.

What are your challenges?

I fully know that there are many more obstacles than the three I listed above – such as legal issues, ROI concerns, job security fears, lack of bandwidth, and more. I’m eager to hear about your challenges or doubts in getting AI adopted across your team or company.

If you are a leader, manager, or supervisor, I’d love to chat with you. You can use this link to schedule a short call.