Google X was building Loon at least a decade ago to solve internet connectivity with laser-networked high-altitude balloons. They clearly benefit structurally from expanding access to the internet to the unconnected population.

So on that front, while I didn’t guess this, it does sound like a good investment.

Furthermore, while I am skeptical about the “DC in space” thing in the short term, and Musk’s timelines are, generously, something like P99.9 outcomes, it seems plausible that in 10 years it might make sense. And at that point Google will want to put their TPUs in orbit, too.

Presumably Starlink for routing throughout space. Google does under-sea cables and domestic internet access too.

They invested in SpaceX long before Starlink was a thibgy

Starlink wasn't working yet, but had been announced as a thing they intended to do.

Ehh, more like Starlink was a thing they said they'd do to justify the investment at the time. Just like they said they'd build an assembly line mass producing Falcon 1's for small launch market, or send a Starship full of artists around the Moon. The first prototypes were three years later.

What are you people arguing about here? This is exactly what is argued here, i.e. that Alphabet is like an investment fund? They had a hunch and it paid off, didn't they? I.e. you're making the point for the argument made?

That 2015 Alphabet was the tail end of the Google X era where the founders were making crazy conviction bets that weee only thinly related to their core area, very different from current Alphabet that really only does make considered strategic investments.

How is that surprising? This is what investors do, don't they?

Having your exponentially more expensive fleet of satellites constantly burning up versus dedicated cables that rarely need replacing

The big use-case for Starlink is typically for people in locations without dedicated cables.

All energy generated on earth contributes to global warming, even solar panels which decrease Earth's albedo. If you care about climate change and the earth you should want the datacenters to be in space.

The earth is limited in the amount of thermal energy it can radiate.

Ah yes, burn thousands of tonnes of methane (assuming Spacex) to launch a data centre satellite into an orbit, then burn thousands more to maintain it and update compute or burn the entire thing in the upper atmosphere, to avoid contributing to global warming by not burning all of that fuel and dry mass.

Both aluminum and iron atoms are heavy and thus make it into the ocean where they both increase the ocean's carbon sink capacity. https://phys.org/news/2021-05-aluminum-affect-climate-ocean-...

From an energy POV, the chemical launch energy required for orbital insertion is 900GJ per satellite (when amortized) whereas the satellite produces 18,900GJ over its 5-yr lifetime, and far more if it lasts longer.

Then we should just outright seed the oceans, half assing it isn't going to do anything.

Uh, that's not how global warming works, you know.

I believe the competent people at Google did a ton of research before any M&A and private investments like SpaceX.

but I doubt they could have thought through every possible outcome or every direction a private investment might take

we humans have limitations. Oftentimes, good investing involves luck and intuition. It may still be far better than gambling, but it can’t be 100% data-driven or explained purely in scientific terms

Who knows? SpaceX could eventually launch something truly groundbreaking or something completely unexpected in the far future, and Google couldn’t have known that when they made the investment. We simply don’t know. Sometimes we just have to attribute it to luck, intuition, or something we can’t fully explain.

One would presume connectivity and services in space as industry and services spread there. Asteroid mining, on orbit manufacturing, space colonies/tourism, data centers in space, centers for working outside traditional government jurisdiction.

starlink cellular connection for android in future

They had a big early investment back when Google was doing more crazy out of scope things.

Similar of SpaceX and Twitter I guess

Google Cloud, low latency cross-continental links, etc.

I don't think gcp customers were hoping for higher latency, 5% packet loss, and modem speeds.

lower latency. You forget about cross-satelite laser links that shorten the latency for cross-continental traffic.

Everything else you said draws from some regular consumer experiences. That's not what I am talking about here.

Modem speeds?

Low cost payloads to space can unlock lots.

    > Data centers in space?

the investment happened 10 years ago

It's telecom infrastructure to increase the number of people who have internet in order to increase the number of people using Google. They might have been considering having satellite internet in Android phones.

YouTube is probably one of the few companies that could benefit from data centers in space.

Rather than co-locating media servers into ISPs you can stick them into space so they're just closer.

Space in NYC isn't cheap. Space in space is free.

> Space in NYC isn't cheap. Space in space is free.

... what?

Say what you want about starship etc etc, but space in space is far from free and will be for a long time.

Who do you pay rent to in space?

Space is free; you don't even pay property taxes.

Transportation costs are not but that's the whole reason you invest into (at the time) a space transportation company.

You can't just get something to space and then leave it there for free. Any maintenance is way more expensive than a data center on the ground (be it via extra trips or extra R&D for autonomous operating).

You can't leave anything on the ground for free either.

Like what is your argument? A new technology is expensive? That's why you invest into bringing the costs down.

Google invested into balloons to share internet (Loom). Which they eventually gave up on but the correct decision is to try new things and then evaluate their future success. Not just armchair decide that change is scary.

I'm not saying space data centers should happen tomorrow but just that ruling them out completely for YouTube is foolish. Like Sports often send the video feed through space back to earth to be processed. You can skip a step there by just sending it to space, processing it there, and then send it back down to subscribers.

If a data centre is in space, its almost always going to be further away than a data centre on Earth.

Most people in the developed world are probably under 500km away from the nearest data centre, with most connections being established over fibre.

A space DC would require wireless communications, over longer distances. The latency is higher. Not to mention, Jitter, Latency and Packet Loss are pretty consequential to video streaming, specifically real time or live broadcasts.

Google has a strong interest in everybody getting better internet (see Fiber, Loon, Fi, etc), which lines up well with Starlink.