For the purposes of this article, I want you to fast forward just a couple of months into 2025. It’s February or March, and o3 is now past safety checks and available. What does that world look like? I’ll post some more personal reflections on o3 a bit later, but I wanted to start thinking about o3 from a product perspective now—it’s not a silver bullet, and it’s not a labor replacement tool in the foreseeable. So enjoy this post from the near future…
Artificial intelligence has sparked more than just technical transformations—it’s reframing how we build and deliver products. With the debut of OpenAI’s O3 model, we move from an era of pure computational might toward a more nuanced view of how AI can drive value in everyday workflows. Businesses are fundamentally information streams, generating and processing “tokens” of data at every turn. The question is: Where in these streams is intelligence most valuable? And how can AI target the tasks that genuinely deliver value to people—both customers and employees—while weaving seamlessly into human workflows?
From Token Streams to Value Streams
To understand how O3 (and AI in general) might reshape our world, it helps to see businesses not just as organizations but as streams of data and interactions—or “token streams.” Every customer inquiry, internal memo, product shipment, and financial transaction is part of this continuous flow. Some tokens in the stream are mundane or “trash tasks” that employees dread; others are high-value choke points where efficiency or precision can make or break the company’s success.
O3’s intelligence becomes a strategic resource for modifying these token streams at the exact points where a slight change could yield outsize returns. Instead of brute-forcing its way across all tasks, O3 can be deployed surgically, tackling the toughest challenges (a complex engineering problem, a large-scale simulation, a risky financial forecast) or automating the repetitive “trash tasks” that bog down employee productivity. Let me open that up a bit:
