If you’re around 40 and suddenly seeing TikToks about ADHD or autism that hit uncomfortably close to home—you’re not alone. A lot of us grew up in an era where these things weren’t caught unless they were severe. We just got labeled “quirky,” “too sensitive,” or “not living up to our potential.”
Why Our Generation Is Figuring This Out Now
Here’s the thing about late-discovered neurodivergence: it reframes everything. Why certain jobs felt impossible. Why friendships were exhausting in ways you couldn’t explain. Why you developed all those coping mechanisms that worked well enough—until they didn’t.
Kids today have language for this stuff. They talk openly about executive dysfunction, sensory overload, and masking. Meanwhile, those of us pushing 40 are quietly Googling “is it normal to rehearse conversations in your head for hours” and realizing… maybe not.
Understanding how your brain actually works isn’t about collecting a label. It’s about finally having context. And once you have that, you can actually do something useful with it—find strategies that work for your brain, learn from others who think similarly, stop blaming yourself for things that were never character flaws.
The Problem With Most AI Conversations About This
I started using ChatGPT to explore the ADHD vs. autism question because, honestly, I wasn’t ready for a formal evaluation. I just wanted to think through it.
But most conversations went sideways. Ask ChatGPT “do I have autism?” and you’ll get either overly cautious disclaimers or vague validation that doesn’t actually help you understand anything. The AI defaults to being supportive, which sounds nice—but supportive isn’t the same as useful.
What I needed was something more like an actual clinical interview: structured, analytical, willing to say “that doesn’t fit” when it didn’t fit.
So I built a prompt that does exactly that.
The Prompt: A Clinician-Style Self-Discovery Interview
This prompt treats ChatGPT as a thinking partner, not a cheerleader. It asks the AI to run a structured interview, reflect back what your answers suggest, and—crucially—be willing to point toward non-autism explanations when the evidence points that way.
Copy and paste this into your favorite LLM
Autism Self-Discovery Interview (Clinician-Style)
I'd like you to act as a clinical psychologist–style interviewer who specializes in adult autism, ADHD, and anxiety differentiation.
I understand you cannot diagnose me, and I'm not looking for a label. What I want is a guided self-discovery interview that helps me understand how my brain works and where autistic traits may or may not apply.
Please run this like a real intake interview:
Ask questions in clear clusters (social intuition, sensory processing, cognition, emotional regulation, developmental history, masking).
After each cluster, reflect back what my answers suggest, including what does and does not point toward autism.
Be explicit about alternative explanations (e.g., ADHD, anxiety, personality, giftedness).
Make falsifiable predictions about my behavior that I can evaluate for accuracy.
Important constraints:
Do not default to validating everything as autism.
Be willing to say "this does not fit autism" when appropriate.
Keep the tone respectful, analytical, and direct — not therapeutic fluff.
Start by explaining the process briefly, then begin with the first question cluster.Why This Prompt Actually Works
Most self-assessment conversations with AI fail because they’re designed to make you feel good, not to help you think clearly. This prompt flips that.
It invites rigorous thinking, not identity affirmation. You’re explicitly asking for analysis, not validation. That changes how the AI responds.
It signals that non-autism outcomes are acceptable. This is huge. By telling the AI upfront that you want it to push back, you bypass the default “be supportive” mode.
It encourages prediction, not just resonance. Instead of “does this sound like me?” you’re asking “what would we expect to see if this were true?” That’s a much more useful question.
It mirrors how actual adult autism evaluations work. Real clinicians look at patterns across domains, consider alternative explanations, and don’t conclude autism just because a few traits are present.
It avoids the “everyone is a little autistic” trap. By asking for specificity and falsifiability, you filter out the vague stuff that applies to basically anyone.
What to Expect
The conversation will feel collaborative and curiosity-driven. You’ll explore patterns across different areas of your life—social situations, sensory experiences, how you think and process information, emotional responses, childhood history, and the ways you may have learned to mask or compensate.
After each section, ChatGPT will reflect back what it’s hearing and what it might mean. Sometimes it’ll point toward autism. Sometimes it’ll say “this looks more like ADHD” or “this could be anxiety” or “this is pretty typical, actually.”
That’s the point. You’re not trying to confirm a hypothesis—you’re trying to understand what’s actually going on.
What I Learned About My Own Brain
When I ran this conversation myself, the conclusion wasn’t autism. It was ADHD plus anxiety—which, honestly, I already knew. I’d been diagnosed with ADHD two years ago and take medication for both conditions. But the conversation clarified why I’d resonated so strongly with autistic experiences in the first place.
Here’s the insight that hit hardest: I share surface traits with autistic adults—feeling different from the social norm, being highly introspective, communicating directly, analyzing people instead of intuiting them when stressed. But the why is completely different.
The AI put it this way: Autism is often “the social world is opaque, so I build rules to understand it.” My pattern is closer to “the social world matters deeply, so my brain scans it for threat and meaning.” Those feel similar from the inside, but they’re opposites at the nervous-system level.
One distinction that stuck with me: “I don’t miss social signals—I sometimes get overwhelmed by them.”
The conversation also made falsifiable predictions about my behavior—things like feeling relief when plans get canceled, rehearsing arguments that never happen, and being emotionally “all in” or oddly detached with little middle ground. Many of it’s predictions landed and I felt validated.
That’s what good self-discovery looks like. Not confirmation of what you hoped to find, but clarity about what’s actually there.
A Few Things to Keep in Mind
This isn’t diagnosis. It’s self-reflection with a structured framework. If the conversation surfaces things that feel significant, consider pursuing a formal evaluation with someone who specializes in adult neurodivergence.
Be honest in your answers. The whole point is understanding yourself better. If you shade your responses to get a particular outcome, you’re just fooling yourself.
Take your time. You don’t have to finish in one sitting. Some of these questions might require real thought, or you might want to observe your own behavior for a few days before answering.
What Comes After Understanding
Self-discovery is step one. But once you have a clearer picture of how your brain works, the next question becomes practical: now what do I do with this information?
If the conversation points toward ADHD—or you already know that’s part of your wiring—AI can do more than help you understand yourself. It can become the external working memory your brain never had. I wrote a full breakdown of how to use AI to actually complete projects with an ADHD brain, covering everything from capturing ideas before they evaporate to staged delivery methods that prevent overwhelm. The techniques there pair well with this kind of self-knowledge.
The Real Value Here
Understanding how your brain works—whether that ends up being autism, ADHD, both, neither, or something else entirely—gives you something powerful: the ability to stop fighting yourself and start working with how you’re wired.