Stop Sitting on the AI Sidelines
AI skepticism is really interesting to me. You wouldn’t know there was any reluctance when you read the daily headlines. Every day a new CEO is touting the benefits and efficiencies of introducing AI to their workforce. Of course, CEOs might be overselling it. AI is the hot buzzword – every executive wants to sound cutting-edge to Wall Street.
However, anecdotally, I’ve had a surprising number of chats with people who haven’t tried using their company’s AI. On the one hand, I get it. The discourse around AI is really confusing and sometimes extreme! Some say AGI will create a dreary dystopia. Others warn of the imminent elimination of countless jobs. Will people automate themselves out of their livelihood? These are valid fears!
Fear paralyzes and invokes apathy. The future has never felt more uncertain. But I am here to tell you that these fears are out of our control. Worrying about them is unproductive. Staying sidelined while this technology blossoms is a mistake. The only way to make the future is to take control of the present: using AI to understand it, and thus shape it. We have to be like Jane Goodall studying chimpanzees. I am inviting you to grab your binoculars and notebook. Let’s talk about how to approach this thing.
Be a Healthy Skeptic
First things first: set your preconceived notions aside. Don’t expect miracles, and don’t expect disaster. Channel your skepticism in a different way. It should be directed in two places simultaneously: outward toward the AI itself, and inward toward yourself. By that I mean, be skeptical of your own intentions!
When interacting with AI ask yourself: "am I seeking validation for something that I want to be true?" If you are, please know that the AI model may try its hardest to do that for you! This is called cherry picking and it’s never good to do. Get into the mindset of finding the full truth–the good, bad, and ugly. Some LLMs protect against cherry picking better than others, but they are all trained on data with inherent bias. So remember to stay skeptical and prompt as a truth seeker, not a cherry picker.
Use the Omniscient Rubber Duck
Now that we’ve discussed the mindset of truth-seeking, let’s move to questioning. When you have an idea, you discuss it with others for feedback, right? This is one reason why networking is so important! Various people to help with various problems! There’s a limitation though: what if no one in your network can help with your unique problem? Well my friend, this is a perfect opportunity to use AI! This is your omniscient rubber duck, trained on the sum total of human knowledge!
Of course, communicating through a chatbox is hard, and you’ll have to provide lots of context. Claude AI tools like the Project feature and Gemini tools like Research Mode will deliver more robust answers. Start with small questions and go from there.
Rubber Duck Examples
The following examples are the start of a two-way dialog. Don't just ask AI something complex and take the first answer at face value.
- You want to architect code a certain way. Explain your approach to the AI in great detail, upload files, and ask it to critique it for positives and drawbacks. Tools like Claude Code and Windsurf are great tools utilizing agentic modes for enhanced context awareness. If it provides code to use, ask about the code.
- You are not a scholar, but you read a research paper and it seems scary. You want more context, so upload the paper to AI and ask it to provide context about the findings. Continue the conversation by asking the AI about its strengths and weaknesses. Tip: enable Research Mode for better analysis!
- Your app crashed and you don’t know where to start debugging. Upload the crashlog to AI and ask it to provide analysis. Again, agentic modes are great for this.
- You want to scale your quality assurance processes but you are running into configuration issues. After you’ve thought about your plan of action, talk to the AI. Upload your problem statement and your proposed solution to Gemini and ask for an assessment. Ask it to poke holes. You might get some duds, but you also might find a gem.
Explore the Unknown
New projects or situations ALWAYS have blind spots that are simply hard to predict! AI won't magically reveal these for you, but it can get the creative juices flowing. Think about it: you have a problem and start by asking the obvious questions first. As you learn more, you question more. As you learn, your thought patterns will change and branch out. Maybe you realize that your original problem is actually a side effect of a much LARGER problem? Or maybe the solution is actually a simple fix, not an arduous project?
Invoke the Socratic Method
Now, let's supercharge your interrogation skills by using the Socratic method. The Socratic method, named after Socrates, is famously used in law school and scholarly debate. The Socratic method dissects ideas through questioning, forcing you to examine every assumption. It’s learning through dialogue!
Sounds amazing, right? Great news: now you don't need to go back to school to use it. You can prompt AI to use the Socratic method, right now.
Open a chatbot and try this simple prompt: I want you to teach me using the Socratic method.
Ask a question about anything. The AI will then ask you questions to clarify your meaning. Think, answer, repeat! Pretty soon you will be thinking about your subject from a totally different vantage point.
Challenge yourself to try to learn anything with it. Try these:
- What causes inflation?
- How do combustion engines work?
Always Play
These concepts will help guide your first steps. Remember: be a truth-seeking examiner of the Socratic school of thought. Don’t be a petty cherry picking validator!
Also keep in mind that play is essential for learning. Make it fun. At your job, start using AI with low stakes tasks: summarizing a document, revising an email, upload documents and ask for analysis. Interrogate it. Try to find its bounds and push the limits.
The first thing I ever asked ChatGPT to do was to flip a coin, tell me the result, and count it. Flip it 5 times. 10 times. 50 times. For every time it lands on heads, multiply the count by 1000. Why? No good reason! I was playing. This kind of play branched out into other curiosities as I developed a sense for the model. It was kind of like playing guitar for the first few times until you get familiar with the fret board!
Try different chatbots! Only use a company approved AI at work. But try the others with your personal questions on your own time! Ask the same questions across a whole bunch of them and pay attention to how they differ in their responses. Anthropic’s Claude is my personal favorite for everything (especially coding). The more you tinker, the more elastic you will be!
Bonus: Playing with Midjourney
Besides chatbots, image generators are another opportunity to explore AI.