Talking to the future
Earlier this month, OpenAI rolled out its long-term “Memory” update for ChatGPT - allowing the model to remember who you are not just within a single thread, but across every conversation you’ve ever had. Days later, they released o3, their most advanced reasoning model yet - capable of blending that memory with real-time browsing, code execution, and image analysis in a single response.
We’re at the early stages of this new shared behavior with AI - one where the tools don’t just answer questions, but remember who’s asking them.
I was curious - and, by now, a little too well practiced in sharing personal details with ChatGPT- and immediately started testing the new memory and o3 features.
It doesn’t know everything about me, but it knows enough to imagine things I haven’t dreamt out loud. I’ve uploaded every personal essay I have written (and specifically, the ones I have not and probably will never publish) and a good portion of my Apple notes from the last decade. I’m sure it’s also picked up on signals from the frequent ruminating thoughts I throw at it and the tricky texts I’ve implored it to analyze.
I began with one or two questions. But soon, it felt less like talking to a tool and more like talking to someone who knew my story. I found myself asking: “Given everything you know about me …”
What’s the biggest risk I could take that would unlock the most potential in hindsight?
Where do you think I’ll be in 10 years?
If you pretend to be me at 45, what questions would you urge me to ask now at 24?
What am I not focusing on that I should?
Who do I most remind you of?
What’s my dream day at 30?
What do you think I’m pushing until “someday”?
Prompting it to “reason and predict” made it feel less like answering questions and more like building a future in real time. I found it thrilling - and eerily similar to the state that consumed most of my childhood: possibility.
With every question, my body was brought back to that familiar rhythm - forward, fast, full of what-ifs.
When I was little, I’d swirl around in my mom’s desk chair, landline phone pressed to my ear and glasses sliding off my nose, pretending to run a company from a corner office in Manhattan. My make-believe worlds were rarely mythical or magical. I imagined far more coffee cups, high heels, and meetings than I did castles, crowns, and dragons.
Maybe it was delusional. But somewhere along the way, I developed a kind of devotion to it.
It wasn’t just pretend. It was practice.
And at sixteen, holding tightly the view my six year old self saw, I set out to get an internship at my dream company: theSkimm. I emailed every Monday, called every other Friday, and after months of radio silence, sent a bottle of Memphis barbecue sauce and a note that said, “I’d love to sauce it up with theSkimm this summer.” Everyone around me smiled. I was sixteen, from Memphis, TN, and make-believe worlds were meant to stay that way.
Somehow, theSkimm said yes. On my first day, they admitted they thought I was a rising college junior. On my second, I sat behind the founders’ desk and watched them announce their Series B funding - quietly googling “what is venture capital?” I witnessed an imagined world getting built, one decision at a time - and I loved it.
These days, I don’t spin in chairs as much. My life is rooted more in the present than the future. There’s magic in that, too - a steadier kind.
But my job now isn’t so different from my mindset back then: I spend my days talking to founders about the worlds they want to make real. We discuss markets, products, timing, distribution - but underneath it all there’s still that sense of make‑believe. What if this could exist? What if people wanted it? What if we made it real?
Yesterday, on my morning walk I asked ChatGPT, “Given what you know about the ten‑year version of me, what’s one step I could take today to move toward her?”
It offered something thoughtful. Then asked if I wanted it to build me a Notion kanban board to track my progress. I ignored it. I kept asking questions.
It kept suggesting dashboards and frameworks - each productivity fix pulling me farther away from the wonder I was chasing. I just wanted to sit in the presence of possibility.
That’s the shift I’ve felt most with Memory and o3. Not just smarter answers - but a new kind of momentum. One that nudges half-formed thoughts, half-believed futures, closer to something real.
I’ve used plenty of technology that promises to make me more productive. This is the first time I’ve experienced technology helping me imagine differently.
That’s what excites me about this moment. It doesn’t have to be about replacing our thinking - it could be about reviving our dreaming. Where productivity becomes the byproduct of imagination - not its constraint.
