Learning AI vs. Doing AI

Hi there,

This week, I spoke at two events. One was about AI in Industry and the other about AI in Social Media. Two very different topics, but they shared one big idea: AI isn’t here just for experts, but for everyone.

The key question is: How do you prepare or "enable" people for this? What's not working is to send everyone back to classroom and make them study hard.

I think it's time to flip the script and spoil the end. Let people access and experience modern AI right away. Introduce more technical knowledge as needed, depending on the user and their intentions.

Let’s jump in!

Talking about hands-on AI, here’s a great opportunity coming up:

In this upcoming hands-on workshop, I'll walk you through 20+ use cases for ChatGPT in data analytics. LIVE on February 7 & 8, 2024. Recording available when you register. (Subscription required, but you can join with a free trial)

In Applied AI, Practice Beats Theory

Fun fact: Even the world’s leading AI experts can't fully explain how AI brains like ChatGPT really work in detail. It’s still an open research question, as OpenAI's Chief Scientist Ilya Sutskever explains in this video:

So as a user, don't worry too much about the intricacies. You can leave that to an armada of AI researchers to figure out. But the good thing is, in the meantime, you can focus on what you can control best:

Find out if and how AI can help you get better at what you already do.

Honestly, I never met a single person who got successful with AI by just reading about it. Yet, many who started with no clue built amazing use cases by jumping into ChatGPT and figuring things out.

And you don’t need an AI degree for that.

There's a great German term for this - "Handlungswissen", best translated as "know-how". It describes practical skills, the so-called "ability" that has been acquired through physical experience and practice.

The enabler: AI accessibility

If you want to learn by doing, then you actually need to be able to do the thing and fail safely.

For example, if I wanted to become a pilot, it’s probably a bad idea to board a plane and "figure it out". Both the cost of failure and the cost of access to the technology are just too high.

Five years or so ago, the only feasible way to get "hands-on" with modern AI was to:

• Get data

• Label data

• Train a model

• Deploy the model

• Find out that your "AI” wasn’t even all that good.

Especially in businesses, that means a high cost of failure because you kept a group of people busy for probably some months.

Today, this has changed fundamentally with multi-purpose AI systems available, as-a-service.

For free.

ChatGPT, Bard, or Bing are just the most popular examples. More niche applications like Magnific AI, Codeium, DeepL Write, or Tome appear by the day. There's literally NO reason why you shouldn't try these out (responsibly, of course - without a critical use case or sharing private data).

And with that, we have the chance to totally change the way we gain knowledge about AI.

"Learning AI Backwards"

Typically, the first rides in an AI services like ChatGPT will be quite exciting. However, this effect will wear off quickly if you're not quite sure how to make the most of it. Chances are, you’ll hit a wall pretty soon where AI answers are kind of "meh" or totally off track.

A lot of people’s AI journey ends right here.

So, you might wonder: Is diving into AI hands-first really the right way to learn it? I would put it like this: Jumping straight into using ChatGPT is the best first step.

Let's flip the script and approach AI from the end, not the start.

Here’s a 3 day action plan:

Day 1: Do something simple

Get ChatGPT Plus and fire up ChatGPT4. But don't sweat the complex stuff yet. Start simple. For example, let it write a recipe or have a critical discussion with you. Go play and have fun - but keep expectations low.

Day 2: Do something more complex that’s worked for others

Before you get too advanced, see what other people have done successfully with ChatGPT. For example, check out this data analytics prompt flow, or build a simple assistant. Try to get more comfortable with the tool and push your own limits a little more.

Day 3: Take a break and reflect

Now, before you jump to any conclusions or enroll in a course, take a moment to reflect. Answer the question: "Could this potentially make a difference in my daily life or work?"

If it's a "no", no stress. Check back soon. But if it's a "yes" or a "maybe", that's your cue to dig a little deeper.

From AI Experience to AI Understanding

If you're still here, it's time to peel back the layers to understand a little more about the basics, the do's and don'ts, and what AI might bring to the table in the future.

So, where do you start? Here are a few breadcrumbs to follow:

• Intro: Elements of AI – One of the world’s most popular AI courses. Non-technical, up-to-date and free by the University of Helsinki.

• Bootcamp: Nocode AI Bootcamp – Taught by IBM's Director of AI, Armand Ruiz, this free, hands-on bootcamp follows the "learning by doing" approach over a 10-week course.

• Comprehensive: The Complete Artificial Intelligence and ChatGPT Course is an in-depth walkthrough of AI applications in business, with a special focus on ChatGPT, taught by AI expert Luka Anicin and business expert Chris Haroun.

If you want to dive even deeper, or build professional AI applications yourself, consider looking at the excellent Machine Learning / AI specializations by Google, Microsoft, or Deeplearning.AI - most of them are free.

Conclusion 

Like any skill, AI proficiency comes from a mix of learning and doing. Today's AI systems make it easier than ever to get your hands on it. Embrace that, but don't stop there. "Experience" is only one component of learning, the other is "understanding the why."

It's about understanding the reasons behind AI's actions, its limitations, and its potential impact on our world. These deeper insights will transform you from a naive user to an informed participant in the AI age.

Think of it like driving a car. Knowing how to operate the vehicle is essential, but understanding the rules of the road and why they exist makes you a skilled and safe driver.

Similarly, in AI, the "how" allows you to use the technology, but the "why" enables you to apply it wisely and creatively in different scenarios.

Stay curious, keep exploring.

See you next Friday!

Tobias