I Interviewed AI—ChatGPT Reveals What It Can (and Can't) Do in 2025

Can an AI truly understand us? Does it have thoughts, creativity, or self-awareness? The movie Her was set in 2025—the year we’re living in now. So I decided to interview artificial intelligence itself. In this fascinating 45-minute conversation, I put OpenAI's ChatGPT on the hot seat to explore what AI is actually capable of today, what it fundamentally lacks, and what the future might hold.

For over two years, ChatGPT has been my daily research assistant, copyeditor, and conversational partner. It’s replaced Google searches, helped me refine complex ideas, and even served as an on-call thought partner for unpacking movies and exploring concepts. With its voice function, conversations feel eerily natural, like talking to an all-knowing assistant who’s always ready with clear, accessible answers.

But I’d never pushed beyond casual use. I hadn’t asked it to explain how it actually works, what it “thinks” (if that’s even the right word), or how it views humanity. So I sat down for a live interview with AI to find out.

What emerged was both fascinating and unsatisfying. ChatGPT is remarkably transparent: it doesn't have emotions, can’t feel happiness or sadness, and acknowledges its creativity is fundamentally limited because it lacks the emotional depth that fuels human imagination. It’s designed with safeguards to be ethical, transparent, and incapable of manipulation or deception. According to the AI itself, fears about unintended consequences are mitigated by its programming.

But hearing an AI explain its own limitations raises deeper questions. What about human error—the kind that led to disasters like Chernobyl? At its current stage, ChatGPT is an incredibly advanced chatbot, but it’s not sentient, self-aware, or capable of true general intelligence. Which made me wonder: what would a conversation with artificial general intelligence, or superintelligence, actually be like? That’ll have to be a future episode.

In this conversation, I explore:

• What AI can and cannot do in 2025, straight from the source

• Why ChatGPT says it lacks emotions, self-awareness, and human-inspired creativity

• How AI views its role in supporting (not replacing) humans

• The built-in ethical safeguards and their limitations

• The difference between advanced chatbots and artificial general intelligence

• What the leap from today’s AI to superintelligence might look like

• Using AI as a tool while understanding what it fundamentally is—and isn't