Kai McKnight

I'm based in Memphis and study finance online at Berkeley College in the City of New York, where I hold a 4.00 GPA. I graduate in August 2027, and after that my focus turns to electrical engineering. I'm working & hoping toward a graduate degree in electrical engineering at Stanford, and I'm building the math and physics foundation to get there now.

Most of what I do outside of class is build things. Sift is an iPhone app I'm developing in SwiftUI. It scans a food barcode, pulls the product's information, and gives it a 0 to 100 health score with a plain explanation of why.

My second project, Datum, is built for civil engineers. When a firm designs a road or a site, the plan set can run hundreds of pages, and one number that doesn't match from page to page can turn into an expensive mistake. Datum is an AI tool I'm building in Python that reads those drawings, compares revisions, checks details like drainage tables against public standards, and flags anything that doesn't line up. The engineer still makes the call. Datum just makes sure they see the problem.

Where I want all of this to lead is SpaceXAI, because I believe it's going to win. Chip production keeps climbing every year, but new power plants and grid connections take years to build. At some point there will be more chips than there is electricity to run them, and power becomes the bottleneck for the entire AI industry. SpaceXAI is the one company built for that moment. On the ground, Colossus already runs at gigawatt scale, and its capacity is valuable enough that even a direct competitor like Anthropic leases compute from it. Beyond the ground, SpaceX is working toward putting compute in orbit, where solar power never stops and the grid no longer matters. No other company pairs frontier AI with its own rockets and satellite network. When power decides who wins in AI, the company that controls the most of it wins, and that is an electrical engineering problem I want to spend my career solving.