AI: A Helpful Tool, But Not a Crystal Ball

If you follow my husband on this social media, you know he's been really passionate about #ArtificialIntelligence. So, for the last few months, I haven't spent one single day without listening to some long talk about AI and its possibilities, challenges, and applications in real life. [laughing]

On one of these, he asked AI for tips on gaining lean body mass. And this was one big example of how AI can't (yet) fully replace human beings, and also the importance of #promptengineering.

The robot told him to consume approximately 3,405 calories per day to promote muscle growth in one-year to achieve his weight goal. If you know a bit about nutrition and bodybuilding, this is the recommendation for highly active athletes. Highly. My husband is NOT it. Well, he is also not sedentary. But, as an atypical father and busy professional, he wouldn't be able to eat all of those calories in a 100% healthy way. Can you see what I mean?

If he takes AI advice without support from a doctor, a nutritionist, or a personal trainer, it wouldn't be a good result. This was ONE small practical example of the dangers of AI. It got me thinking...

Well... We could say that is ALL about the prompt. If he had told the AI more about his routine and habits, it could've made more assertive suggestions. At the same time, a GOOD health professional wouldn't prescribe a diet like this without asking more questions. In this tool, AI is supposed to give an answer even without knowing all the context.?

What is the conclusion? It's not about ONE conclusion, but some warnings and #lessonslearned.

FIRST LESSON: Without a smart prompt, AI could be mistaken.?

It is programmed to give you an answer based on your question. So, you have to know not only WHAT, but HOW to ask.?

That's an interesting point since "prompt engineering" may become one of the new professions or required skills AI creates for the future.

(So it could not only make some professions useless but also create new ones. Maybe not in the same proportion).

SECOND: AI regular tools are not ready to know when the recommendation needs more information to be precise. BUT, it could become.

This is just the beginning. After all, one of the scariest (and most incredible) parts of AI is its ability to learn and evolve by itself.?Also, we're just having a taste of what has already been developed now.

Since the first AI tool public launch 5 months ago, new updates and other tools have been launched. FIVE months!

THIRD: Once we have already given the world access to many kinds of AI tools, it may get into the hands of naive people that may not know how to ask the question and may wrongly trust the answer it gets from an incomplete prompt.

This topic may deserve a whole discussion on it and it's the main reason I decided to write this article (after months of AI private classes at home, haha ??).

We are here learning and questioning the AI tools. We are being critical of it. This is good, and will also help them to evolve.

But some people may use free (or more accessible) AI tools to save money on proper support, like health. (Especially in countries where health care is expensive). We must ensure people will be educated on this.?

Is it an incredible tool to create content and brainstorm ideas? YES.?Is it inevitable to grow and become part of our everyday lives? Almost sure it is. Should we blindly trust it? Absolutely NOT.?

Isn't Artificial Intelligence programmed to think like Humans? Can we blindly trust Humans? We all know we can't. So, it's logical to also not 100% trust AI. Let's be sure to emphasize this when we are praising its amazing abilities. Deal?

Written by: Andressa Pereira on May 1st (Labor Day in Brazil).

Choosing the title for this article

Can you guess who I asked for help when choosing the title for this article? Who could think like a Human but be impartial like a robot? Yeah, AI.

I asked the OpenAI tool to give me catchy and attention-grabbing title suggestions for this article. I told #chatGPT what was it about, and it gave me a few suggestions. It was not easy to choose, here are some of the other suggestions I rejected, but liked it:

1. "The Pitfalls of Blindly Trusting AI: What You Need to Know"
2. "Don't Be Fooled: AI Can't Replace Human Judgment (Yet)"
3. "Why AI is Not (Yet) a Replacement for Human Intelligence"
4. "The Human Side of AI: Why We Still Need to be in the Driver's Seat".
5. "The Fine Line Between AI and Human Judgment: Why You Need Both"
6. "AI and Human Intelligence: Finding the Sweet Spot"
7. "AI and the Need for Critical Thinking: Lessons Learned"
8. "Why AI Needs a Reality Check: Lessons from the Field"
9. "The Risks of Blindly Trusting AI: What You Need to Know"
10. "Why AI is Not a Magic Bullet: Lessons from the Trenches".

Which one would you choose?