6 Months Post-Grad
6 months ago, I joined Cerebras Systems. Today, we went public on the NASDAQ.
97% of deeptech hardware startups die. Even more, today's semiconductor companies are entrenched by an oligopoly of a few. Yet Cerebras' 10-year forecast on the future of artificial intelligence, along with 10 years of persistent R&D innovations— despite near-death experiences— with the Wafer-Scale Engine, cultivated to the inflection point: effectively, the new age of AI. One fast enough where inference speed is becoming the standard, while models leapfrog each other on SOTA benchmarks like a game of tag: the moment you're ahead, someone's already coming for you.
Having spent most of my time at Stanford pondering about international security, tech was never on my radar— let alone something as deep tech as Cerebras. Instead, I spent most of my time oscillating between whether I'd be a good fit in foreign policy or should rather stick to the corporate route of consulting, and the seemingly inevitable path to law school.
In an increasingly techno-political world, I've come to believe that chips and compute are the new oil (thank you Chris Miller for your rigorous research and superb writing), steering international security and driving what will come down in history as the most rapidly developed technology. As such, deep technological knowledge across all layers of the artificial intelligence stack: energy sources, data center operations, semiconductor architectures, hardware supply chains, model capabilities, frontier AI research, and even the surface layer of B2B SaaS use cases will be a superpower.
Certainly, for when the reverberations of AI permeate every institutional foundation— governance, security, economy— those who wield deep technological literacy will be the ones shaping generations to come.
And of course, selfishly, I've always been greedy for knowledge. At Cerebras, where innovations are pioneered across every layer of the AI stack, that greed finds no shortage of fuel.
Finally, learning by doing as an operator has led me down many rabbit holes, but has solidified confidence in my ability, judgement, and execution. The dichotomy of immense pressure yet profound freedom.
Indubitably, though, I'll take the weight of the former over the comfort of the latter. I simply cannot live a boring life.
Today, I stand at the precipice of what will be decades of technological transformation, unpredictable, exhilarating, and at times, deeply unsettling. Ever so beautiful as much as the ugly.
And we must be ready for that.
