I often think of ChatGPT today the same way many of us regarded the iPhone a decade ago. Back then, people loved to point out that Android offered more customization and richer features. Those observations were true, but it didn’t stop the iPhone from defining the entire smartphone experience for billions of users. Similarly, although other AI models may excel in certain tasks—scoring higher in coding benchmarks or offering unique capabilities—ChatGPT has already captured the public’s mindshare. It’s become synonymous with AI in much the same way iPhone once became shorthand for smartphones.

Some people question whether the transformer architecture at the heart of ChatGPT can ever truly become “intelligent.” To me, this critique overlooks how emergence works. One ant alone can’t achieve much, but a colony of thousands manifests collective intelligence. Likewise, a single neuron in the human brain is limited, yet billions of them working together create consciousness. The exact upper bound of neural network capabilities remains unclear, largely because we still don’t fully grasp how these systems work at scale. History suggests that simple elements, combined effectively, can unlock higher-level abilities we never saw coming.

Others debate whether artificial general intelligence (AGI) is even possible. In my view, AGI is often defined in philosophical or sci-fi terms that may not match the technology we’re building. Instead of focusing on some hypothetical “free will” in machines, I see AI as an immensely powerful tool, one that’s already extremely useful and bound to get better. As it evolves, it will open up entirely new application spaces, just like the introduction of smartphones did for mobile computing.

It’s also worth noting that our understanding of AI won’t stand still. Researchers are currently exploring alternatives to the transformer architecture, and with the substantial influx of talent, funding, and computing power into the AI sector, we can expect a steady stream of incremental improvements, punctuated by occasional breakthrough or disruptive innovations. That’s why I’m optimistic about AI in general, and OpenAI in particular. Ten years from now, we might look back on today’s ChatGPT and reflect on how it paved the way for the transformation of computing—just as the iPhone sparked the smartphone revolution.