Q&A: Social Capital chief on the cost of spectrum

Chamath Palihapitiya surprised everyone in 2015 when he announced he was forming a company and prepared to spend billions on spectrum. In the end, although he registered for the auction, he ended up not bidding.

Why this matters: Palihapitiya and his team are plotting a different approach to shaking up the wireless by rejecting "traditional spectrum" and instead using "ultra low-cost satellites." In an interview, Palihapitiya talked with Axios about what changed his mind about spectrum bidding and offered some hints at what he's doing instead.

Axios: You considered bidding in the spectrum auction - but decided not to, right?

Palihapitiya: We could have spent 2,3,4,5 billion. Instead we will spend 1/10th of the cost and have 100 times the impact. It would have been more splashy to just spend the money but I think this will be more legitimate in the long term.

What are you doing instead?

Wireless is a big, complicated problem and you can't hit it with just one solution. We have a major coverage challenge, which some people are trying to solve with balloons and drones, but we think there's a better way to do this via ultra low-cost satellites. There's also the software layer, which addresses everything happening on the device. And then there's the bowels of it all, the OSS/BSS systems (the operating and billing systems used by cell service providers), which in today's world of all-IP traffic should just happen in the cloud. We are innovating in each of these areas. The part we still need to figure out is the last-mile relay infrastructure, to the extent it's needed, which today is pretty misunderstood. But no matter what, all this technology advancement adds up to not needing traditional spectrum and opens the door to use unlicensed and lightly licensed spectrum which, as you can imagine, is meaningfully cheaper and more abundant.

Traditional satellites are pretty expensive — on the order of hundreds of millions of dollars to build and launch, right?

We think there is a lower cost way to tackle the coverage issue with a very different kind of satellite constellation.

When will this start showing up?

I think you can expect small proofs of concept this year with a full constellation within 24-36 months.

One of your companies, LotusFlare, is already working on modernizing the software cell phone service providers use?

LotusFlare already has some serious traction building software for the device as well as a cloud-based OSS/BSS layer. The other pieces are early but very promising. We just need to bring all these parts together. In a few years, we believe there could be a complete cloud replacement of today's carriers.

Doesn't most of the world have cell service or is in the process of getting it?

You'd be surprised how much of the world still isn't connected. We believe everyone in the world, anywhere in the world, should have affordable internet access. Additionally, you can't just think about current needs. There are all these remote devices today that don't have access but want it. Think about agriculture and shipping as two examples. All these industries are getting smarter and our systems today are totally unprepared for our near-future coverage needs.