The Costly AI Skills Mistake Most Companies Are Making—and How to Fix It

At the time of writing this (2024), it's hard not to talk about AI.

There’s a very important lesson here with the current pace of AI adoption, and the lack of meaningful skills to support it.

AI, particularly generative AI (which are not the same, fyi), has opened a transformational shift in how we work, learn, and interact with the world. Yet, as with any major technological shift, successful ROI doesn’t happen overnight.

The journey from curious “hobbyist” to confident “adopter” is a gradual one, and I cannot overstate how much patience you need in developing here.

Social media doom-scrolling makes it easy to feel pressured to learn everything about AI instantly.

Everyone and their dog is an AI expert today, and apparently ‘they’ can make you ‘master AI in 7 days’. Be wary of these people, they will stunt your chance of success long-term.

Building a deep understanding of such a transformative technology requires time and effort.

And to be quite frank, no one has mastered it yet. They probably never will as it’s always evolving.

You know my views on this already.

Meaningful AI adoption is more than just knowing how the tools work. It's about cultivating a mindset and building the behaviours that allow us to integrate AI meaningfully and responsibly into what we do.

→ And that takes time.

The 3 Stages of AI Literacy: Hobbyists, Experimenters, and Adopters

There are so many bloody maturity models out there right now.

While mine is not as fancy as a consulting firm, I believe it’s simple to use.

My work these last few years has shown most people are navigating through three broad stages of AI skills maturity: hobbyists, experimenters, and adopters.

Let's unpack these:

Hobbyists

Hobbyists are those who dabble in AI, experimenting with tools like ChatGPT in their personal time but haven't yet applied it systematically in their work.

They're curious, but they haven't reached a level of skill where AI significantly impacts their productivity. Mostly they create cat pictures and get AI to write crap social media posts stuffed full of emojis.

Experimenters

Experimenters have begun incorporating AI into their daily tasks, testing out its capabilities, and exploring use cases in real-world contexts. They're still in the learning phase, figuring out what works, what doesn’t, and how AI fits into their broader workflow.

I like this level the most. To experiment, fail and learn is a beautiful thing. The majority of people who play here will do very well.

Adopters

Adopters have fully embraced AI, using it effectively and strategically in their context to enhance work.

They’ve developed a level of comfort and expertise that allows them to apply AI in ways that generate meaningful, long-term value.

Moving from one stage to the next is a slow process. Often frustratingly slow in a world where we expect immediate results.

That’s totally fine. It’s a necessary progression.

Without taking the time to fully understand the nuances of AI and how it can be harnessed, you risk missing out on the true potential of the technology.

Always get clear on the ‘what, why and how’.

Classic advice for a reason.

Be intentional with AI skill building

This will sound counterintuitive, and yes CEO of x company, I know you want the ‘AI Effect’ today.

But with AI literacy, going slower, or shall we say being more intentional, can reap rewards for years - perhaps even decades.

I’ve seen this in some of my work with clients.

Senior executives have crazy expectations for workers to become ‘AI Experts’. They don’t even know what that means.

If we’re talking about tools like ChatGPT, becoming an expert on that with its almost daily updates is like chasing after your 5-year-old when they see an ice cream truck fly by.

Solid fundamentals will help, no doubt.

But fundamentals don't = fully capable expert.

AI is not static.

Learning the fundamentals and taking time to put them into practice is key. Yes, I know that's hard in a world where you need more than 1 week to show 'ROI'.

By encouraging a more deliberate approach, companies can craft the mindsets, new behaviours, and technical, and human skills to navigate AI transformations at large.

I know I’m preaching to the choir here.

(Note: Being more deliberate with crafting AI skills does not mean building bloated 3-month + learning experiences. No one wants or needs this!).

80% of AI projects fail because of this

Another report I’m reading, in what I must say, is an era for ungodly amounts of reporting on one topic, focuses on the root causes of failure for AI projects.

If I’m being fair, the findings of these failures apply to L&D projects too (more on that in the Steal These Thoughts newsletter).

Anyway, one of the biggest factors for failure was being given the time for a project to succeed. You see executives are drinking the kool-aid. They think that what needs at least a year to succeed can be done in a week.

The writing is on the wall for most projects before they start.

You have no doubt suffered this exact problem with countless L&D projects.

Think of all the projects that have died because:

• Expectations were unchecked
• A problem was not defined to solve
• The resources you need to succeed weren't provided
• You were given 1 week when you need 1 year

One word to define this - misalignment.

AI literacy is about building a long-term capability, not a short-term fix.

For a workforce that is not just technically competent, but equipped with the critical thinking, creativity, and adaptability needed to succeed in an AI-driven future.

Final thoughts

As a good BCG article once told me, "Treat Gen AI upskilling as a marathon, not a sprint".

Yes, you need to move fast to help people unlock the potential of new technology. But, you also have to be smart. People won't just get it after some 30-minute online course.

They will need more hand-holding than you think, and you need to inject a dose of realism into the 'time to become proficient' with your AI tools of choice. Marathons are a mixture of both fast and slower-paced elements.

The investment in Gen AI fundamentals at most companies is criminally low.

Don't fall into the trap of tools before educating on the basics. I've seen this back-fire too many times. As the wise Uncle Ben said, "With great power, comes great responsibility" - and too many are forgetting the final part of that famous quote.

As I said in a recent, Steal These Thoughts! Newsletter:

With all-time high levels of use across millions of Gen AI tools and all-time low levels of AI literacy, we could be heading for a skills car crash of our own design.

Too many forget that AI is only as good as the human using it.

It's, perhaps, the greatest ‘mistake’ made in all this AI excitement.

Here's five things I suggest you do:

1. Teach AI Fundamentals: What is AI and Gen AI, and what is not? How LLMs work, etc

2. Behaviours + mindset: How to think critically and validate outputs. Understand AI hallucinations. Know when and when not to rely on AI tools

3. Practical use cases: Not cat pics, real work impact. You could combine this with tools for experimenting.

4. Picking the right tools: Not every AI tool is created equal, so know the opps and limitations of yours

5. Upgrade human skills: You won’t go far without a strong sense (and clarity) of thinking and analytical judgment.

The key to all of this is time, patience and intention to build the right skills.

Sometimes that will be fast, others it will be slow.

In sum: Don't make the mistake of rushing the process of crafting meaningful AI skills and behaviours.

📌 This was an extract from my weekly, Steal These Thoughts! newsletter. Join 4,000 L&D pros to improve learning and performance with the power of modern technology.

📖 Read more

The Hidden Impact of AI on Your Skills

How To Get More People Adopting AI at Work To Build Modern Skills

How To Design Meaningful AI Skills Programmes