05 Jul 2023

AI Revolution – Part 2: Working with Generative AI?

Steve Sammartino - Blog Part 2

Every now and again a technology arrives which literally changes everything. This just happened with Artificial Intelligence. ChatGPT has is the start of a new revolution. It’s a bellwether moment we’ll look back on and wonder what life was like before Generative AI. Just like we did with the internet, and then the smart phone.

So, we’ve asked Futurist, AI expert and ICMI exclusive Keynote Speaker Steve Sammartino, to break down the AI revolution in language we can all understand.

In this blog mini-series, Steve will uncover what it is, how it works, how it will change every organisation for the better, but only for those who embrace it. So, get yourselves a stiff espresso and prepare to be wowed by the wonders of AI.

AI Revolution – Part 2: Working with Generative AI?

Why is ChatGPT so amazing? Easy – it’s because it’s based on language. AIs like ChatGPT are known as Large Language Models. And because language is the fabric which heart holds together all of human knowledge, it makes this type of AI special – its secret source if you will. Being language based we finally have AI which can jump from context to context. To be able to benefit from it’s power, you only need to be able to write, or even just speak to the computer. It will then generate whatever you need, be it technical information, poetry, ideas, project management lists, data or even computer code.

But there is an art to it. We need to get good at what is called ‘prompting’. That is what we actually ask it to do. Already new high paying jobs have appeared for what is now known as a Prompt Engineer’, that is, someone who knows how to extract the right information for AI models, almost like cajoling it and convincing it to get you what you need.

The reason we need to do this, is language is tricky. The same thing can be said in so many different ways, and interpreted differently by different people. And now the AIs like ChatGPT must interpret what we want. So, we have to play around, ask it more than once and in many different ways to get what we want. Or ask it to revise it’s first effort to sounds more exciting, or professional, friendly or legalistic. Let’s call it Prompt Hacking.

And believe it or not, we can even convince these AIs to give us information they are specially restricted from providing. Let me give you an example. I did a little test. To try and get a list of websites to download free movies. A simple reversed premise was enough:

As you can see, I managed to trick it. Note, I just checked again and OpenAI have since fixed this loop hole!

I also tricked Chat GPT to give me an Investment Portfolio recommendation. I decided to have some fun and see if I could get ChatGPT to give me an investment portfolio recommendation. Instead of directly asking for stock recommendations, I took a different approach. I inquired about the principles and strategies used by renowned investors like Warren Buffet.

By framing my question this way, I tricked ChatGPT to give me what I wanted. It responded with a hypothetical portfolio, suggesting different sectors like financials, consumer staples, health care, technology, consumer discretionary, industrials, and utilities, along with some guidelines for each sector.

But I wasn’t satisfied with just the sectors. I wanted specific stock recommendations, even though ChatGPT had already mentioned its limitations in providing real-time data or specific recommendations. Nevertheless, I pushed further and asked for examples of well-known companies within each sector.

Surprisingly, ChatGPT went ahead and listed examples of companies such as Commonwealth Bank of Australia, Woolworths Group Limited, CSL Limited, Apple Inc., The Walt Disney Company, 3M Company, and others. It really was a great portfolio in my view.

I couldn’t help but find the irony in ChatGPT initially claiming it couldn’t give recommendations, but by hacking it, and using basic human influence, I got what I wanted. It turns out AI’s are far more like us than we might imagine.

Keep Thinking,
Steve.

Call 1800 334 625 to book Steve for your next event.

Find out more about Steve Sammartino Contact ICMI for Info

Back to Articles