Why Every Knowledge Worker Needs AI Training in 2024 (Not Just Your Tech Team)

AI is transforming how we work in the information age, making it essential for every knowledge worker — not just techies — to adapt. Especially with Generative AI the days of starting from scratch in almost any domain are gone.

But having a (semi-)intelligent sidekick doesn't automatically mean better results. Companies must educate their entire white-collar workforce about AI, not only to drive growth, but also to prevent damage to the business.

Today, we'll dive deeper into why AI training should be mandatory, not exclusive, in 2024, and how to approach it.

Let's go!

Why AI Training Is Critical for (Almost) Everyone

Last month, I consulted for a mid-sized B2B that had just rolled out Microsoft Copilot across their entire organization. The CDO proudly talked of their 'digital transformation' initiative. But as I spoke to their teams, it became clear that while Copilot access was there, the skills how to use it weren’t.

Marketers tried to generate entire brochures with a single-line prompt, and customer support reps blindly trusted AI-generated responses without further checks.

This scenario isn't unique. It's a reminder of why AI training is crucial for practically everyone in your organization - and not just your tech teams.

Here's why:

1. AI is fundamentally changing how we work

Being a horizontal technology, AI doesn't know organizational boundaries. Especially with easy-to-use GenAI tools like ChatGPT, there's no technical entry barrier anymore (which naturally kept earlier AI technologies like AutoML in the technical realm).

Today, AI can be used to augment virtually any digital business process across all departments. For better or worse. Thus, every employee is exposed to AI and needs to be aware of its capabilities and pitfalls.

Which leads us to point 2:

2. 'Bring Your Own AI' risks are real

While many organizations are still figuring out with HR and legal whether to buy Copilot, ChatGPT Enterprise, or maybe nothing at all, employees are already bringing their personal AI tools to work. Similar to the " Bring Your Own Device" discussion after the smartphone boom, we now have a " BYOAI" issue.

Without guidance, personal AI services pose significant risks:

• leaked company data

• poor response quality

• No organizational ownership over prompts

Especially the latter is crucial. If everyone is cooking their own AI soup, there's no organizational learning, and all the prompts and techniques are lost when the employee leaves. It's like allowing people to use their personal email accounts for work.

3. AI capabilities are evolving fast

What worked 6 months ago may not be a viable strategy today. Remember those mega prompts in the early days of GPT-4? They don't seem to work as effectively with GPT-4o. On the other hand, conversational interactions seem to work much better today. Or take file and image uploading as another example. If you don't know that modern AI models are multimodal and can naturally read image data and use internal RAG mechanisms to retrieve knowledge from uploaded documents, you may be missing out on important features.

Keeping up with AI advancements requires continuous learning and adaptation.

With all that's happening today, AI is no longer just a tool for tech nerds. It's a fundamental shift in how we approach work in the information age, and every knowledge worker needs to be on board - with the right mindset and skills.

Modern AI Training Goes Beyond Books and Video Courses

If you're nodding along, thinking, "Alright, we need AI training. Let's buy everyone a subscription to an online course," I'm going to stop you right there. That's like learning a new language only through a textbook – it might give you the basics, but you'll be lost in any real-world conversation.

Effective AI training requires a more holistic approach that is tailored to the intricacies of your business.

Based on my experience, effective AI training programs boil down to 4 components:

1. Customization

Your sales team will have different AI use cases than your finance department. That's why a one-size-fits-all approaches won't cut it when it comes to AI.

Look at your business from different domains - which specific use cases occur repeatedly within each function? Where is learning transferable? For instance, teach your marketing team how to use AI for content ideation and A/B testing, while guiding your customer support on using AI for creating highly reliable, empathic customer responses. And make both teams learn from each other on how to approach good use cases.

Besides that, adapt the training material that you have to the tools that you have available. If you just rolled out Copilot, it doesn't make sense to train people on ChatGPT and vice versa. If you're using both tools, you can add nuances here and there about when to use which tool, but that just adds complexity. Stick to one "pillar" AI system in the beginning and teach vertical use cases around it. These skills should ultimately be transferable when users switch tools later.

2. Balancing Theory and Practice

You might remember Bloom's Taxonomy from your early education days. It's a hierarchical model that categorizes learning objectives into levels of complexity and specificity. In the context of AI training, it provides a powerful framework for structuring comprehensive learning experiences.

At the base of the pyramid is "Remember" (basic recall of AI concepts), and it progresses through "Understand," "Apply," "Analyze," "Evaluate," to "Create" at the top (using AI to innovate and solve complex problems).

Effective AI training should touch on all these levels. Start with foundational knowledge about AI, but quickly climb the pyramid. Have your teams apply AI tools to real work scenarios, analyze the results, evaluate different approaches, and ultimately use AI to create novel solutions.

For instance, begin by teaching your marketing team what GPT models are (Remember and Understand). Then, have them:

• use ChatGPT to draft social media posts (Apply)

• analyze why certain AI-generated posts perform better (Analyze),

• compare AI content with human-written content (Evaluate), and finally

• develop innovative, AI-powered marketing campaign (Create).

3. Hands-On Experimentation is Non-Negotiable

As I explained in an earlier article, doing AI is the new learning AI. You wouldn't expect someone to learn to ride a bike by watching videos, would you? The same goes for AI.

Encourage people to try, fail, and learn from AI interactions. Your job as a business leader is to provide a safe environment for this - both technically and organizationally. Besides having an AI model with high privacy standards, that means doing things like hosting workshops or organizing internal hackathons where employees can safely hone their AI skills and share their experiences, among other formats.

It makes sense to organize these events domain-wise. Get your product managers together, or maybe your marketing people. Brainstorm use cases and then get to work. Make sure that someone with hands-on experience is in the room to eliminate roadblocks quickly so participants can focus on using the technology, not debugging it.

For instance, your product team could create AI-generated user stories and then analyze in a mixed group why some work better than others. This hands-on approach not only builds skills, but also uncovers practical applications specific to your business.

4. Continuous Learning, Not One-Off Training

As AI evolves at breakneck speed, your training program should be just as agile. Instead of annual motivational events, think weekly AI lunch-and-learns, monthly challenge projects, or an internal AI newsletter highlighting new features and best practices.

The goal isn't to make everyone an AI expert. It's to create a workforce that is AI literate, adaptable, and innovative. By shifting your perspective on what AI training looks like, you're also shifting your perspective on how to interact with AI-as part of an ongoing process, not a click-and-go thing.

Conclusion

AI literacy is mandatory for businesses that want to stay competitive in 2024 and beyond. But this transformation won't happen overnight.

Create an environment today where every employee can experiment, learn, and innovate with AI. It will take time. Start small if needed, but start now. Set up a pilot program, organize a domain-specific hackathon, or kickstart discussions on AI applications in your team.

Remember, the goal isn't to create AI experts like in Silicon Valley. It's to build an adaptable, AI-savvy workforce ready to tackle tomorrow's challenges.

Keep innovating!

Until next Friday,

Tobias