3 ms·

> At Google, where I worked for a few years, it was interesting to see how Google's understanding of search (publicly disclosed), and the infrastructure to host

> At Google, where I worked for a few years, it was interesting to see how Google's understanding of search (publicly disclosed), and the infrastructure to host it (kept secret) kept it comfortably ahead of competitors until the design space was exhausted. At which point Google stopped moving forward and everyone else asymptotically approached their level of understanding and mastery. This is the "markets mature and commodify over time" thing. What companies are supposed to do in those cases are one of two things. One, keep investing the money into the market or related ones so you keep having an advantage. Or two, if there is nothing relevant and adjacent to productively invest in, return it to shareholders as dividends or share buybacks so they can invest it in some other unrelated market. But space seems like it would be the first one big time because of the amount of stuff that still has yet to be developed. Starlink was an obvious example of something in that nature, and then it's going to be things like "put datacenters in orbit so you can use solar without worrying about clouds or nighttime" and "build robots that can do semi-autonomous work in places far enough away for both human presence and round trip latency to be an inconvenience" etc. We'd be living in Star Trek by the time they'd run out of something more to do.

palata 1y ago
> We'd be living in Star Trek by the time they'd run out of something more to do. Chances are that we will be living on the consequence of the end of fossil fuels and the rise of climate change long before that, though.
- AnthonyMouse 1y ago
  The end of fossil fuels seem pretty boring. The worst case is that you replace them with nuclear, which has costs on par with fossil fuels to begin with. If we're lucky some kind of cheap scalable energy storage tech is developed and then energy costs less than it did historically.
  - palata 1y ago
    > The worst case is that you replace them with nuclear Not remotely the worst case. How do you expect to power all the ships needed for globalisation with nuclear power? What about planes? Can a rocket take off with a nuclear engine? Fossil fuels account for 80% of the energy we use, electricity merely 20%. A whole lot of those 80% come from use-cases that were built around fossil fuels (how do you make plastic and all the materials that depend on it with nuclear energy?), and we don't (yet) have a way to replace that with something else. Try to power a ship with electricity... Hydrogen, you say? We would need a lot more energy to produce enough hydrogen to replace oil. So we're going from "we don't have a way to compensate for the lack of oil" with "our solution is to not only compensate, but actually produce more energy than what oil was giving us". All that while currently living a mass extinction and having already missed the reasonable objectives for global warming. So we have a few decades to get there, and what we have seen in the last few decades is that all we have achieved is making the problem worse.
    - zizee 1y ago
      There are many nuclear powered ships. If the decision is between nuclear proliferation, or the end of the global economy, I'm betting we'll choose the former. If you have enough electricity, you can manufacture avgas, methane and whatever other fuels you need for aerospace.
      - palata 1y ago
        > There are many nuclear powered ships What's the ratio of nuclear-powered big ships vs non-nuclear-powered big ships? > If you have enough electricity We're talking about moving our current electricity production entirely out of fossil fuels (because we produce a lot of electricity with them), then multiplying that production by 5, and at this point we're only producing the same amount of energy as today. But of course that's not enough, because we then need to use a lot of that energy to produce what's needed to replace oil, e.g. hydrogen. We currently need what... 10-15 years to build a nuclear power plant? We're talking about building multiple orders of magnitudes more of them in a few decades, together with the electrical network and of course everything that needs to be re-engineered now that they can't run with oil anymore. And we're currently using oil for a reason: it's super dense, there is nothing more convenient. And what value does it add? Nothing. It's just for replacing what currently works. Who will pay for that? Where will the money come from? And this has to be done in a context where geopolitical instability will grow every year (because it is a fact: our access to abundant fossil fuel is coming to an end; Europe has seen it since 2007). And of course in a context where we are not remotely thinking about doing it. In the last decades, we as a society have actually kept accelerating in the opposite direction. How realistic do you think your scenario is, really?
- mensetmanusman 1y ago
  Population is crashing too fast for that to happen.
overfeed 1y ago
> put datacenters in orbit so you can use solar without worrying about clouds or nighttime Data centers in space are a pipedream until we have a material-science breakthrough: radiating heat into space is too inefficient at rack-scale, let alone DCs.

palata 1y ago

> We'd be living in Star Trek by the time they'd run out of something more to do. Chances are that we will be living on the consequence of the end of fossil fuels and the rise of climate change long before that, though.

AnthonyMouse 1y ago
The end of fossil fuels seem pretty boring. The worst case is that you replace them with nuclear, which has costs on par with fossil fuels to begin with. If we're lucky some kind of cheap scalable energy storage tech is developed and then energy costs less than it did historically.
- palata 1y ago
  > The worst case is that you replace them with nuclear Not remotely the worst case. How do you expect to power all the ships needed for globalisation with nuclear power? What about planes? Can a rocket take off with a nuclear engine? Fossil fuels account for 80% of the energy we use, electricity merely 20%. A whole lot of those 80% come from use-cases that were built around fossil fuels (how do you make plastic and all the materials that depend on it with nuclear energy?), and we don't (yet) have a way to replace that with something else. Try to power a ship with electricity... Hydrogen, you say? We would need a lot more energy to produce enough hydrogen to replace oil. So we're going from "we don't have a way to compensate for the lack of oil" with "our solution is to not only compensate, but actually produce more energy than what oil was giving us". All that while currently living a mass extinction and having already missed the reasonable objectives for global warming. So we have a few decades to get there, and what we have seen in the last few decades is that all we have achieved is making the problem worse.
  - zizee 1y ago
    There are many nuclear powered ships. If the decision is between nuclear proliferation, or the end of the global economy, I'm betting we'll choose the former. If you have enough electricity, you can manufacture avgas, methane and whatever other fuels you need for aerospace.
    - palata 1y ago
      > There are many nuclear powered ships What's the ratio of nuclear-powered big ships vs non-nuclear-powered big ships? > If you have enough electricity We're talking about moving our current electricity production entirely out of fossil fuels (because we produce a lot of electricity with them), then multiplying that production by 5, and at this point we're only producing the same amount of energy as today. But of course that's not enough, because we then need to use a lot of that energy to produce what's needed to replace oil, e.g. hydrogen. We currently need what... 10-15 years to build a nuclear power plant? We're talking about building multiple orders of magnitudes more of them in a few decades, together with the electrical network and of course everything that needs to be re-engineered now that they can't run with oil anymore. And we're currently using oil for a reason: it's super dense, there is nothing more convenient. And what value does it add? Nothing. It's just for replacing what currently works. Who will pay for that? Where will the money come from? And this has to be done in a context where geopolitical instability will grow every year (because it is a fact: our access to abundant fossil fuel is coming to an end; Europe has seen it since 2007). And of course in a context where we are not remotely thinking about doing it. In the last decades, we as a society have actually kept accelerating in the opposite direction. How realistic do you think your scenario is, really?
mensetmanusman 1y ago
Population is crashing too fast for that to happen.

overfeed 1y ago

> put datacenters in orbit so you can use solar without worrying about clouds or nighttime Data centers in space are a pipedream until we have a material-science breakthrough: radiating heat into space is too inefficient at rack-scale, let alone DCs.