So, why artificial intelligence?
As far as we know, the brain may be the only thing in the universe capable of studying itself. Humans are curious creatures, and it's in our nature to question everything, even our own capacity to question. So maybe it's not surprising that intelligence, once you start looking for it, seems to show up everywhere. I mean, we live in a world where trees warn each other of danger through underground networks, bees communicate through dance, ants build sprawling colonies, and elephants mourn over the bones of their dead. Intelligence is remarkably ubiquitous, and AI has become the most compelling framework we have for studying it. Rather than simply observing intelligent behavior, it enables us to build systems that reproduce it.
I blame my neuroscience background, but it's this perspective I hold that draws me from studying intelligence to trying to build it. But the deeper I've gotten into AI, the more I've realized that every breakthrough seems to expose another gap in our understanding. We have language models that outperform the average human on many benchmarks, yet they still struggle with arithmetic or spelling. The same models can produce remarkably introspective writing about consciousness yet our deepest intuitions insist that they're not conscious at all. And perhaps most strikingly, they consume enormous amounts of energy to answer questions that the human brain solves on just about twenty watts.
Building the future of intelligence on this planet means sitting with these exciting contradictions. What do we mean by understanding? By reasoning? By learning? By sentience? Building increasingly capable systems has, somewhat paradoxically, become one of the fastest ways to discover how incomplete our theories of intelligence really are. it convinces me that the next leap forward won' come from simply scaling today's ideas, but from developing a deeper understanding of intelligence itself. I want to help build AI that earns its place by solving real problems and extending human capabilities in meaningful ways, not by treating massive datasets and compute budgets as an indefinite cure-all. It's as humbling as it is exciting, and I don't think I'll run out of things to be curious about anytime soon.
With that motivation out of the way, below is a descending timeline of the milestones that best illustrate my journey so far.