5 ms·

I recently heard the idea that the price of energy is going to start approaching zero from Chamath Palihapitiya on Lex Fridman's podcast. It hit me as a big, al

I recently heard the idea that the price of energy is going to start approaching zero from Chamath Palihapitiya on Lex Fridman's podcast. It hit me as a big, all-changing thing. If fusion is really solved, we can probably mass-produce it within years if not months, so we should start thinking of the implications now

colechristensen 4y ago
The price of energy has always been approaching zero. Restated, the amount of energy available to an individual has generally grown exponentially, with pauses here and there because civilization. We might be resuming that growth in the next few decades. Fusion isn’t and won’t be solved in the immediate future. There is a long path towards commercial viability and a lot of milestones along the way… and not really a great method of associating them in a popular science sort of way that doesn’t seem like “we’ve solved fusion!” Real significant progress is indeed being made and is expected to continue, people are going to be excited then disappointed by milestones being announced because we’re closer but not there. Fusion isn’t just always thirty years down the road though, we can see the destination getting closer.
- EZ-Cheeze 4y ago
  I think all progress will accelerate significantly with the dawn of true AI. The fusion scientist can now ask ChatGPT for ten (or 10,000) ideas on how to solve a particular roadblock, and a lot of the suggestions will be actually good. The cost of intelligence has suddenly become near-zero, and that will supercharge everything to do with innovation
  - 988747 4y ago
    LOL, no. Asking ChatGPT is no different than just googling the question, except the answer is more nicely formatted. ChatGPT has no actual intelligence in it.
    - EZ-Cheeze 4y ago
      It can synthesize things into new things. Eureka moments on tap
      - oneoff786 4y ago
        So can a six year old.
  - anotherman554 4y ago
    "The fusion scientist can now ask ChatGPT for ten (or 10,000)" Since the ChatGPT can't perform empirical experiments and update it's suggestions based on those experiments it has no way to tell a "bad" idea from a "good" one.
bryanlarsen 4y ago
In many jurisdictions the cost of delivering a kilowatt-hour is already more than the cost of producing it. So cheap energy is not going to have a substantial impact on normal people unless we switch to wireless electricity delivery or something.
- EZ-Cheeze 4y ago
  Not just cheap though - near zero. If we have an infinite tap of it, we can afford to waste a lot of it in the process of delivering it. The inefficient becomes efficient when you plug in infinities and zeroes
  - bryanlarsen 4y ago
    Even if fusion heat is free, it still costs significant amounts of money to convert heat into electricity. Most of the cost of a coal plant is converting heat into electricity, and the last coal plant built in the US cost $600M to build. Energy costs might drop an order of magnitude, which is significant, but isn't going to cause any sort of systemic change.
  - Brometheus 4y ago
    That's called god parity See: https://reneweconomy.com.au/why-god-parity-will-be-the-end-of-centralised-generation-51446/ https://reneweconomy.com.au/why-god-parity-will-be-the-end-o...
  - oneoff786 4y ago
    That would probably increase the costs of delivering it
curiousgal 4y ago
Who's we? If fusion is solved, the country that has will just become a net energy exporter (somehow, if it's the US, idk how they would ship that power). Nothing much will change for the rest of the world, they will just switch from one dependency to the other.
- stepbeek 4y ago
  Energy transport over long distances is difficult and inefficient iirc. While sone energy storage ships well (like hydrogen) it’s unlikely that there’ll be intercontinental electrical lines. Not an electrical engineer so happy to be told I’m wrong.
- EZ-Cheeze 4y ago
  Interesting point that I hadn't considered. I just automatically assume that when fusion gets invented, it will be shared with the whole world so we can all partake in da utopia... but maybe it will be hoarded instead :(
  - chasd00 4y ago
    The technology will inevitably spread and deals will be made. I don’t think it would be hoarded any more than fission power was/is hoarded.
- defrost 4y ago
  Australia plans to export solar power in two forms; short haul to Singapore via an undersea cable, long haul to Germany via cracked Hydrogen reformed as Ammonia (Larger molecule, easier to handle, harmless Nitrogen byproduct).
Brometheus 4y ago
Don't get your hopes up yet: This just demonstrated that the released energy (in form of heat etc) is higher than the energy input – yet the released energy hadn't been converted to electricity that could have powered the lasers. Supposedly this would be done by a heat engine, with can normally extract 60% of energy input. So 2.5 * 0.6 = 1.5, so it is yet again not producing any surplus energy.
- EZ-Cheeze 4y ago
  We'll know more on Tuesday!!
- WithinReason 4y ago
  More importantly, laser efficiency is around 1%. At "scientific breakeven" you create more energy than the laser delivers. To make it a viable energy source you need to produce at least 100x more energy from a single pulse than scientific breakeven.
  - tuatoru 4y ago
    And there are storage and ohmic losses in the power supplies for the lasers. It's inefficiencies all the way down.
simiones 4y ago
You (or maybe Chamath Palihapitiya) are making a very very wrong assumption that a working fusion power plant would produce free energy. Nothing could be further from the truth: with any half-way plausible design today, fusion power will be the most expensive form of power generation known to man. While some of the fuel is nearly free (deuterium and to some extent even the tritium), it's more realistic to think of the reactor itself as the fuel - since the reactor will be under constant very high energy neutron bombardment, it won't last for more than a few years, maybe a decade at best. And since a fusion energy reactor will be one of the most highly advanced engineering accomplishments in human history (it requires extraordinary structural resistance to hold up the necessary equipment in the presence of the gigantic magnetic fields, extraordinarily advanced magnets, a blanket of liquid lithium and boron that gets pumped continuously and somehow recycled etc.), having to rebuild one every 10 years will be extraordinarily expensive (and we should expect something closer to 2-5 years for the initial designs, even assuming no explosions). In contrast, a fission reactor is already extremely expensive, but has similar output to any contemplated fusion plant, has been proven to be able to be continuously operated for >50 years once built, and requires only slightly more expensive fuel (uranium is really not that expensive compared to its energy density - certainly nowhere near as expensive as gas or coal). Edit: oh, and this seems to be related to a breakthrough at the NIH, presumably on their inertial-confinement fusion experiment, which (despite some rosy rhetoric) is a weapons research program, with no realistic possibility of ever being transformed into a fusion power plant.
- EZ-Cheeze 4y ago
  I think he was talking about decentralized grids & renewable. Let's talk about fission - I used to say I'd build 70,000 nuclear reactors if I was in charge of a rich country. Would I have enough fuel to power them all? Your third paragraph painted a pretty vivid picture in my head, good writing.
  - tuatoru 4y ago
    Would you have enough water to cool them all?
  - 4y ago
Fatnino 4y ago
The same things [1] were said about fission 70 years ago. Didn't happen. [1] https://matt-rickard.ghost.io/content/images/2022/01/image-15.png https://matt-rickard.ghost.io/content/images/2022/01/image-1...

colechristensen 4y ago

The price of energy has always been approaching zero. Restated, the amount of energy available to an individual has generally grown exponentially, with pauses here and there because civilization. We might be resuming that growth in the next few decades. Fusion isn’t and won’t be solved in the immediate future. There is a long path towards commercial viability and a lot of milestones along the way… and not really a great method of associating them in a popular science sort of way that doesn’t seem like “we’ve solved fusion!” Real significant progress is indeed being made and is expected to continue, people are going to be excited then disappointed by milestones being announced because we’re closer but not there. Fusion isn’t just always thirty years down the road though, we can see the destination getting closer.

EZ-Cheeze 4y ago
I think all progress will accelerate significantly with the dawn of true AI. The fusion scientist can now ask ChatGPT for ten (or 10,000) ideas on how to solve a particular roadblock, and a lot of the suggestions will be actually good. The cost of intelligence has suddenly become near-zero, and that will supercharge everything to do with innovation
- 988747 4y ago
  LOL, no. Asking ChatGPT is no different than just googling the question, except the answer is more nicely formatted. ChatGPT has no actual intelligence in it.
  - EZ-Cheeze 4y ago
    It can synthesize things into new things. Eureka moments on tap
    - oneoff786 4y ago
      So can a six year old.
- anotherman554 4y ago
  "The fusion scientist can now ask ChatGPT for ten (or 10,000)" Since the ChatGPT can't perform empirical experiments and update it's suggestions based on those experiments it has no way to tell a "bad" idea from a "good" one.

bryanlarsen 4y ago

In many jurisdictions the cost of delivering a kilowatt-hour is already more than the cost of producing it. So cheap energy is not going to have a substantial impact on normal people unless we switch to wireless electricity delivery or something.

EZ-Cheeze 4y ago
Not just cheap though - near zero. If we have an infinite tap of it, we can afford to waste a lot of it in the process of delivering it. The inefficient becomes efficient when you plug in infinities and zeroes
- bryanlarsen 4y ago
  Even if fusion heat is free, it still costs significant amounts of money to convert heat into electricity. Most of the cost of a coal plant is converting heat into electricity, and the last coal plant built in the US cost $600M to build. Energy costs might drop an order of magnitude, which is significant, but isn't going to cause any sort of systemic change.
- Brometheus 4y ago
  That's called god parity See: https://reneweconomy.com.au/why-god-parity-will-be-the-end-of-centralised-generation-51446/ https://reneweconomy.com.au/why-god-parity-will-be-the-end-o...
- oneoff786 4y ago
  That would probably increase the costs of delivering it

curiousgal 4y ago

Who's we? If fusion is solved, the country that has will just become a net energy exporter (somehow, if it's the US, idk how they would ship that power). Nothing much will change for the rest of the world, they will just switch from one dependency to the other.

stepbeek 4y ago
Energy transport over long distances is difficult and inefficient iirc. While sone energy storage ships well (like hydrogen) it’s unlikely that there’ll be intercontinental electrical lines. Not an electrical engineer so happy to be told I’m wrong.
EZ-Cheeze 4y ago
Interesting point that I hadn't considered. I just automatically assume that when fusion gets invented, it will be shared with the whole world so we can all partake in da utopia... but maybe it will be hoarded instead :(
- chasd00 4y ago
  The technology will inevitably spread and deals will be made. I don’t think it would be hoarded any more than fission power was/is hoarded.
defrost 4y ago
Australia plans to export solar power in two forms; short haul to Singapore via an undersea cable, long haul to Germany via cracked Hydrogen reformed as Ammonia (Larger molecule, easier to handle, harmless Nitrogen byproduct).

Brometheus 4y ago

Don't get your hopes up yet: This just demonstrated that the released energy (in form of heat etc) is higher than the energy input – yet the released energy hadn't been converted to electricity that could have powered the lasers. Supposedly this would be done by a heat engine, with can normally extract 60% of energy input. So 2.5 * 0.6 = 1.5, so it is yet again not producing any surplus energy.

EZ-Cheeze 4y ago
We'll know more on Tuesday!!
WithinReason 4y ago
More importantly, laser efficiency is around 1%. At "scientific breakeven" you create more energy than the laser delivers. To make it a viable energy source you need to produce at least 100x more energy from a single pulse than scientific breakeven.
- tuatoru 4y ago
  And there are storage and ohmic losses in the power supplies for the lasers. It's inefficiencies all the way down.

simiones 4y ago

You (or maybe Chamath Palihapitiya) are making a very very wrong assumption that a working fusion power plant would produce free energy. Nothing could be further from the truth: with any half-way plausible design today, fusion power will be the most expensive form of power generation known to man. While some of the fuel is nearly free (deuterium and to some extent even the tritium), it's more realistic to think of the reactor itself as the fuel - since the reactor will be under constant very high energy neutron bombardment, it won't last for more than a few years, maybe a decade at best. And since a fusion energy reactor will be one of the most highly advanced engineering accomplishments in human history (it requires extraordinary structural resistance to hold up the necessary equipment in the presence of the gigantic magnetic fields, extraordinarily advanced magnets, a blanket of liquid lithium and boron that gets pumped continuously and somehow recycled etc.), having to rebuild one every 10 years will be extraordinarily expensive (and we should expect something closer to 2-5 years for the initial designs, even assuming no explosions). In contrast, a fission reactor is already extremely expensive, but has similar output to any contemplated fusion plant, has been proven to be able to be continuously operated for >50 years once built, and requires only slightly more expensive fuel (uranium is really not that expensive compared to its energy density - certainly nowhere near as expensive as gas or coal). Edit: oh, and this seems to be related to a breakthrough at the NIH, presumably on their inertial-confinement fusion experiment, which (despite some rosy rhetoric) is a weapons research program, with no realistic possibility of ever being transformed into a fusion power plant.

EZ-Cheeze 4y ago
I think he was talking about decentralized grids & renewable. Let's talk about fission - I used to say I'd build 70,000 nuclear reactors if I was in charge of a rich country. Would I have enough fuel to power them all? Your third paragraph painted a pretty vivid picture in my head, good writing.
- tuatoru 4y ago
  Would you have enough water to cool them all?
- 4y ago

Fatnino 4y ago

The same things [1] were said about fission 70 years ago. Didn't happen. [1] https://matt-rickard.ghost.io/content/images/2022/01/image-15.png https://matt-rickard.ghost.io/content/images/2022/01/image-1...