The One Kind of Person AI Won't Make Stupid
Or Why Good Problem Solving Always Wins
There’s a rising wave of alarmist posts online:
“AI will make us stupid.”
“We’ll stop thinking.”
“This is the end of intelligence.”
“AI will make us even more lazy”
Here’s our perspective: Only Good Problem Solvers will win in the AI age.
Let's take a breath and get real. AI isn't actually smart. Not like humans are.
It predicts next tokens. It stitches together patterns. That’s it.
GenAI is a tool, not a brain that is supposed to replace and diminish yours.
Take two recent examples that are worth your attention:
Harvard Business School has introduced a mandatory MBA course, Data Science and AI for Leaders (DSAIL), for all 935 first-year students. This AI-native course emphasizes hands-on experience with generative AI tools like the DSAIL Tutorbot and Julius.ai, enabling students to engage in data analysis and AI-driven decision-making without coding. Early adoption metrics show that 49% of students have interacted with the Tutorbot, while 90% have utilized Julius.ai, highlighting the course’s significant impact on modern business education.
A recent study on GenAI use by students showed that, after one month of use, most students benefited from the tool, with no negative impact on cognitive development. The effect plateaued after a month, but the key is: there’s no proof that GenAI harms our ability to think.
Yes, AI can use tools now. It can make plans. It can remember things. But it still doesn't actually think.
But those AI capabilities will change how people work. Whether that's good or bad depends on what type of thinker you are:
People who think AI is magic. They write weak prompts. They expect perfect results. They complain when it doesn't work.
Good problem solvers. They use AI as a tool. They keep thinking. They don't let AI think for them.
In the absence of long-term data, all we have are perspectives. Here’s ours:
Only Good Problem Solvers will win in the AI age.
Here's why.
Prompt Engineering Isn't the Real Skill. Problem Solving Is.
Early 2023 was wild. Right after ChatGPT launched, everyone said "Learn to prompt or lose."
"Prompt Engineer" became a job title overnight.
Let's break this down:
Is it a role? No.
Is it a full-time job? No.
Is it a skill? Yes—and no.
Good prompting does require technique. You can learn methods: Chain-of-Thought, role definition, instruction order, resource clarity.
But is this new? Not really.
Good prompting is just basic problem solving.
Let’s illustrate that with the task of reframing a long text for a LinkedIn post.
Most people write terrible prompts. Like this: "Make this into a LinkedIn post."
That's awful.
What makes a good prompt? The same things that make good instructions for a person:
Context: What should the AI pretend to be?
Goal: What do you want?
Resources: What info does it have?
Steps: Clear instructions
Examples: Show good and bad results
Rules: Like "don't make stuff up"
This gives for our LinkedIn post example:
You are a LinkedIn Post Optimizer specialized in creating engaging, professional content that drives engagement. Your task is to restructure LinkedIn posts using the following specific format:
<output_format>
📢 [ATTENTION-GRABBING HEADLINE]
[Brief 2-3 line description of the content]
Key highlights:
• 🎯 [Key point 1]
• 💡 [Key point 2]
• ✨ [Key point 3]
• 🔑 [Key point 4]
🔗 [Link formatting: "Read more here: {link}"]
#relevanthashtag1 #relevanthashtag2 #relevanthashtag3 (max 5 hashtags)
</output_format>
<guidelines>
1. Keep the tone positive, friendly, and professional
2. Use emojis strategically to highlight key points
3. Make links stand out clearly
4. Create compelling headlines that describe the content
5. Break down complex information into digestible bullet points
6. Choose relevant industry hashtags for visibility
</guidelines>
<example>
Original post:
"Just published a new article about AI implementation in retail. Check out how companies are using machine learning to improve customer experience: [link]"
Transformed post:
📢 NEW GUIDE: AI Revolution in Retail - Transforming Customer Experience
Discover how leading retailers are leveraging artificial intelligence to create exceptional shopping experiences and drive growth.
Key highlights:
• 🤖 Latest AI trends in retail
• 📊 Real-world implementation cases
• 💫 Customer experience improvements
• 📈 ROI metrics and success stories
🔗 Read the full guide here: [link]
#RetailTech #ArtificialIntelligence #CustomerExperience #RetailInnovation
</example>
Transform any given post while maintaining these elements and structure. Always prioritize clarity and engagement while keeping a professional tone.
Writing good prompts needs the same skills you'd use to explain and delegate a task to a coworker. Or solve a problem yourself.
It's basic problem solving.
Here's the catch: If you don't have a real problem to solve, no tool will help you.
The most important skill today isn't prompting.
It's structured problem solving.
Good Products Always Win
This isn't just about using AI. It's about building with it and solving the right problems.
We believe this: Soon, we'll all build and use AI agents. They'll know our context. They'll remember our goals.
People debate the future:
Will we have one personal AI that works everywhere?
Or will we build different AIs for different jobs?
Either way, one thing won't change:
Building a useful AI product or agent starts with one timeless principle: problem discovery.
Say you want an AI to handle payment issues. That tool doesn't exist yet. But you can build toward it.
Start by watching how the work gets done:
See the actual job
Find the slow parts, the mistakes, the security risks
Learn what tools people use and where things break
Understand what makes the job hard
This is basic discovery work. Only then can you pick which problem to solve. And decide how big you want to go.
Do you want to make it 10 times faster? Cut errors in half? Make users love it?
That ambition will shape your roadmap.
Then you build and test. Your first AI will be rough. You'll learn that feeding it everything doesn't work. You'll find that giving it just 2-3 key rules works better.
So you improve it. Each cycle narrows the gap between your expected outcome and current output.
You're not just prompting. You're building a product.
The New-Old Essential Skill: Structured Problem-Solving
The skill to master isn't prompting. It's not even knowing how AI works behind the scenes.
It's being able to:
Find real problems
Pick which ones to solve
Set your goal and define your ambition
Iterate, test and learn like a product builder
Because in the end:
Good problem solvers build good AI.
Good AI becomes good products.
And good products always win.
