Advice for non-technical people on building their AI knowledge

The Golden Question

A question I’m always asked online is: “If I’m not a technical person, what should I learn when it comes to AI?”.

In times of rapid change, we naturally want to make sure that we’re keeping our knowledge and skills at pace. While AI might seem like some dark magic only reserved for the most technical people, I believe that everyone has something to learn.

Of course, if your goal is to become a Machine Learning Researcher or Engineer, you have to build technical skills.

But if you’re a leader, product manager, UX designer, marketer, or any other non-technical person, there’s fundamental knowledge and skills that can serve you for the rest of your career.

I studied Human Factors and AI Engineering in university, so my recommendations will be based on what I learned back then and what I’m learning/doing now.

AI isn’t going anywhere and it’ll slowly become integrated in all aspects of our life and work over the next few years. So now’s the best time to invest an hour a day to learn and set yourself up for success in the future.

This plan will help you:

1. Gain a strong understanding of how AI systems are trained, how they work, and how they generate outputs.
2. Strengthen your AI literacy to have more productive and thoughtful conversations during your day-to-day and at work.
3. Build your intuition of what AI can or can’t do, so you can apply it in your own life or within your company’s workflows, products, services, or industry.
4. Become more effective at using AI tools.

Nothing in life is to be feared, it is only to be understood.

– Marie Curie

Step 1: Start with first principles and fundamentals

Most of your questions (and fears) about AI can be answered by understanding the fundamentals. Think of it as building the base of your knowledge pyramid.

If you don’t spend enough time building the base, your knowledge will be shaky at best and you’ll have an incomplete picture of this transformational technology.

I’ve also realized that AI systems become much less scary and mysterious once you understand how they work, even if you’re not building your own.

For this, I recommend three free courses that you can take:

1. AI For Everyone by Andrew Ng
2. Making Friends with Machine Learning by Cassie Kozyrkov
3. Generative AI Learning Path by Google Cloud

There’s some overlap between AI For Everyone and Making Friends with Machine Learning, but the latter goes deeper into the maths and algorithms behind AI systems. I’ve done all three courses and found them all useful in their own ways.

If you spend 1 hour each day on these, it’ll take you about two weeks to finish them (unless you wanna binge them like a season of Ted Lasso!).

Step 2: Dive deep into real-world applications of AI

Now that you built the base of your AI knowledge pyramid, you’ll gain deeper insight into the real-world applications of AI. In the past few months, we’ve been exposed to hundreds of headlines of “Company X is using AI to do Y”.

Without understanding the fundamentals, we gloss over the details and don’t ask the right questions. If we’re looking to harness the power of AI in our life and work, we have to start building our intuition and the ability to analyze specific use cases:

• What are the specific ways AI is being applied in an industry or a specific workflow?
• Did this company train their own AI system or use an out-of-the-box solution from a provider like OpenAI?
• Did this company fine-tune an existing model on their proprietary data to increase performance and build a competitive advantage?
• Are they measuring the AI’s performance for those specific tasks? How is it performing?
• How do they monitor or maintain the performance of their AI system over time?
• How do they test their systems for safety to minimize undesirable outputs? Is there a human in the loop?
• How are their customers reacting to their applications? Do they love it? Do they hate it? Is it solving their pains and needs?
• Was this an appropriate use of AI or can a similar (or better) solution be implemented without it?

Since I started writing Year 2049 almost two years ago, I’ve shared plenty of case studies about how companies are deploying AI within their products and services. Check out the case studies page to see the entire list.

Step 3: Get your hands dirty and have some fun (please!)

This might be the easiest (and most fun) step that not enough people do.

You have to spend time, experiment, and play with all these available AI tools. Everyone knows about ChatGPT, but I’m still surprised when people tell me they never used it or only used it a handful of times. There are plenty of other tools if ChatGPT isn’t your cup of tea.

Personally, I like going on Product Hunt and browsing through all the new tools that are released every day.

The more you interact and engage with AI systems, the more you’ll appreciate their capabilities and recognize their weaknesses.

So, open up ChatGPT, Midjourney, or any other AI tool every day and have fun. Don’t let the blank page intimidate you. Experiment. Give them random tasks. Modify your prompts and see how the output differs each time.

One experiment I had a lot of fun with recently was building a custom chatbot following the 1-hour Chat with Your Data course from DeepLearning.AI. What I loved about the course is that the video lesson is playing on one side, and an interactive code editor is on the other. All the code is pre-written and you just need to run it. It helped me understand the step-by-step process of building a chatbot based on a custom set of documents.

If you’re looking for inspiration, YouTube is a great place to look at how others are experimenting with AI tools.

A sneak peek of what’s coming to Year 2049

This is by no means an extensive guide to everything you need to learn, but a good starting point. There are more things to learn based on your specific job or industry, and I’m working hard on new content to help with Steps #2 and #3 specifically.

This will include:

• More case studies for specific jobs (e.g. Designer, Teacher, etc.) or industries (e.g. Education, Banking, Real Estate, Climate Change Relief, etc.).
• More practical tutorials for AI tools like Midjourney, Tome, Code Interpreter, and much more.

I’m a one-man team for the moment and this takes a lot of work, so I appreciate your patience and support.

‍