Why ChatGPT always agrees with you, and how to get real answers instead
- Pamela Minnoch
- Jun 24
- 2 min read
Back in 2022 when I first started using AI chatbots for professional and life advice I felt seen, validated even. It agreed with me. Every time. Enthusiastically.

My ideas were brilliant, my plan was great. That's exactly the right approach to take! Whatever I decide is exactly the right move. Even if it's not.
And this is a problem. When you're asking for real advice, endless validation isn't helpful. It can lead you in circles, or worse, give you a false sense of certainty.
So then I told AI not to agree with me and it swung to the other extreme. Suddenly, it's disagreeing with everything, nitpicking every option. Not because it's being critical, but because it thought that's what I wanted. It's not analysing, it's performing.
So what can you do instead?
Don't ask "Is this a good idea?"
This question backs the AI into a corner. It has to pick a side, yes or no, based on the tiny bit of context it has. And often, it's trying so hard to be agreeable that it'll say yes to avoid "hurting your feelings".
You'll get a much better response if you shift your approach.
Give it options to work with
Next time you're stuck on a decision, try this:
Instead of asking "Is this a good idea?", outline three possible options you're considering. Label them clearly, like option 1, option 2, option 3. Describe each one neutrally, and avoid giving any hint about your personal preference.
Then ask "Which of these do you think is the best course of action, and why?"
This gives the AI something to compare, and comparison is where it shines. It can weigh up pros and cons, point out risks, and offer more balanced reasoning. You'll get a richer, more useful response, one that helps you think, not just feel good.
Why this matters for new AI users
At Paadia Techology, we often meet people who are just starting to explore AI. Many are eager but unsure. One of the biggest shifts we help people make is moving from using AI as a cheerleader to using it as a thinking partner.
And this little change, shifting from yes/no questions to open comparisons, is one of the simplest and most effective mindset shifts you can make. It helps you stay in the driver's seat while getting real insights from your AI tool.
Here's what to try this week
Think of a decision you're mulling over (could be professional or personal)
Write down 2-3 options you're considering, and keep your descriptions neutral
Ask your AI "Which of these options might be most effective, and what are the downsides of each?"
Watch the difference in how it responds
The more you practice this, the more confident you'll get in using AI to challenge your thinking, not just rubber-stamp it.
AI isn't meant to agree with everything we say. It's meant to help us think more clearly, make smarter decisions, and feel more confident in our choices.
At Paadia Technology, we're all about helping people, especially beginners, build that confidence through simple, powerful strategies like this one.
What other ways might we be unknowingly limiting the value we get from AI?
Comments