10 ms·

They built a device around as big as a golden retriever, shipped it to a planet 130 million kilo meters away, and that device is generating its own electricity

They built a device around as big as a golden retriever, shipped it to a planet 130 million kilo meters away, and that device is generating its own electricity and generating as much O2 as a tree. I don’t know man, I think that’s pretty advanced technology. We went from room sized transistors to nm sized ones in less than a century. No way we will need a billion years to catch up.

iancmceachern 4y ago
I love the use of a golden retriever as a unit of size. I have poodles personally, but they come in various sizes so it's a bit less clear.
- rzzzt 4y ago
  It needs to be golden so that corrosion doesn't alter its dimensions.
babypuncher 4y ago
The first transistor was about the size of a Raspberry Pi but I think your point still stands.
- gnu8 4y ago
  That might be an interesting project, configuring a Raspberry Pi to emulate a single transistor and then using them to build a computer.
  - babypuncher 4y ago
    Building something useful out of that would end up comically large. I should probably qualify my "first transistor" statement; the device I was referring to is the first prototype created at Bell Labs in 1947. The first transistor that someone could actually buy to build a circuit out of came out a year later and was only a little larger than today's through-hole mounted transistors. They did get used to build early minicomputers, before integrated circuits took over.
    - geenew 4y ago
      >> That might be an interesting project, configuring a Raspberry Pi to emulate a single transistor and then using them to build a computer. > Building something useful out of that would end up comically large. The transistors are MicroSD-size, not RPi-size, but still, it has been done: https://www.megaprocessor.com/ https://www.megaprocessor.com/
unity1001 4y ago
> They built a device around as big as a golden retriever, shipped it to a planet 130 million kilo meters away, and that device is generating its own electricity and generating as much O2 as a tree Which could be easily accomplished by specialized bacteria dispensed to that planet's atmosphere. Yeah. Technology is billions of years behind.
- antasvara 4y ago
  I'm curious: do we currently have bacteria that could survive in the atmosphere of Mars? My understanding is that it would require controlled conditions and monitoring to make bacterial growth feasible under those conditions.
  - lazide 4y ago
    Conditions that support Liquid water is only transitory there. So you’re generally correct. Most extremophiles we’re aware of tend towards high temps or if low temps, in the context of extreme high salinity and water presence. Which don’t line up with Mars much. We’d have to be managing the environment they grew in, which makes it hard. The underlying issue of course being energy gradients and biochemical availability of that energy. Life ‘eats’ to survive, but if the only energy gradients are feeble and biochemically hard to access, it’s not a good environment for life as we know it.
- alpaca128 4y ago
  The difficult part about that would be undispensing the specialized bacteria once we figured out that it backfired and contaminated a whole planet.
  - baruch 4y ago
    One you send humans that you can't sterilize before flight you will inevitably introduce bacteria into the environment anyway, and you will not be preselecting them either.
  - peddling-brink 4y ago
cercatrova 4y ago
Room sized transistors? Lol. Then how big were the computers that linked them together?
blue_dragon 4y ago
> On Perseverance we use up to 300 watts to produce about 8 grams per hour of oxygen, which represents not much more than 10 percent efficiency relative to the amount of electrochemical power it actually takes to pull apart the CO2 molecule. ... It’ll still require a reliable power source, of course. According to Dr. Hecht, a nuclear reactor capable of around 10 kilowatts (currently being developed by NASA) should do the trick. To me, this is not "generating its own electricity". The device is still an incredible accomplishment, even more so because it's operating on Mars. But it doesn't generate its own electricity. They still need to connect it to a separate energy input to run the reaction. I want to clarify this just in case anybody skimming the comments gets the impression that chemically splitting the CO2 into CO and oxygen gas somehow produces enough energy to sustain the reaction without external input.
cmrdporcupine 4y ago
Multiple decades of very slow progress in fusion, space travel -- and I dunno, almost everything else -- should be enough to make it clear that Moore's Law doesn't apply to every other technological endeavour. There's no reason to assume exponential type improvements. That's not how the world works.
- psadri 4y ago
  First we build AGI, then we unleash it on all these other problems.
- JumpCrisscross 4y ago
  > no reason to assume exponential type improvements Did we read the same article? This was a test. The full-scale version “would produce oxygen at a rate equivalent to several hundred trees” with close to an order of magnitude more efficiency “because a bigger machine can run at lower pressures, saving energy on compression.” There are better reasons to expect exponential improvements than not.
- ascar 4y ago
  Transistor progress has been especially astonishing, but if you just look at the progress in basically all areas of living in the last 200, 1000, 5000 years there is absolutely no reason we won't see incredible further improvements in the next century. They don't have to be exponential doubling every 18 months. Steady incremental improvements compound just fine over decades. I just don't see any of your opening statement being even close to correct given the context. There was even significant progress in space travel and fusion in the last 50 years (e.g. first reactors being net positive after startup and reusable rockets with commercial space flights).
- Retric 4y ago
  The deeper point isn’t to compare mores law to all new technology but rather to consider how long it takes to get close to some fundamental limit. Early engines had horrible efficiency but that rapidly increased over time up to a reasonable fraction of hypothetical limits. You see this all over the place, modern guns aren’t multiple orders of magnitude improvements over guns built 100 years ago. But they are orders of magnitude better than the absolute earliest guns ever built. So, when looking at a prototype the question isn’t can it be improved but by how much can it be improved.

iancmceachern 4y ago

I love the use of a golden retriever as a unit of size. I have poodles personally, but they come in various sizes so it's a bit less clear.

rzzzt 4y ago
It needs to be golden so that corrosion doesn't alter its dimensions.

babypuncher 4y ago

The first transistor was about the size of a Raspberry Pi but I think your point still stands.

gnu8 4y ago
That might be an interesting project, configuring a Raspberry Pi to emulate a single transistor and then using them to build a computer.
- babypuncher 4y ago
  Building something useful out of that would end up comically large. I should probably qualify my "first transistor" statement; the device I was referring to is the first prototype created at Bell Labs in 1947. The first transistor that someone could actually buy to build a circuit out of came out a year later and was only a little larger than today's through-hole mounted transistors. They did get used to build early minicomputers, before integrated circuits took over.
  - geenew 4y ago
    >> That might be an interesting project, configuring a Raspberry Pi to emulate a single transistor and then using them to build a computer. > Building something useful out of that would end up comically large. The transistors are MicroSD-size, not RPi-size, but still, it has been done: https://www.megaprocessor.com/ https://www.megaprocessor.com/

unity1001 4y ago

> They built a device around as big as a golden retriever, shipped it to a planet 130 million kilo meters away, and that device is generating its own electricity and generating as much O2 as a tree Which could be easily accomplished by specialized bacteria dispensed to that planet's atmosphere. Yeah. Technology is billions of years behind.

antasvara 4y ago
I'm curious: do we currently have bacteria that could survive in the atmosphere of Mars? My understanding is that it would require controlled conditions and monitoring to make bacterial growth feasible under those conditions.
- lazide 4y ago
  Conditions that support Liquid water is only transitory there. So you’re generally correct. Most extremophiles we’re aware of tend towards high temps or if low temps, in the context of extreme high salinity and water presence. Which don’t line up with Mars much. We’d have to be managing the environment they grew in, which makes it hard. The underlying issue of course being energy gradients and biochemical availability of that energy. Life ‘eats’ to survive, but if the only energy gradients are feeble and biochemically hard to access, it’s not a good environment for life as we know it.
alpaca128 4y ago
The difficult part about that would be undispensing the specialized bacteria once we figured out that it backfired and contaminated a whole planet.
- baruch 4y ago
  One you send humans that you can't sterilize before flight you will inevitably introduce bacteria into the environment anyway, and you will not be preselecting them either.
- peddling-brink 4y ago

cercatrova 4y ago

Room sized transistors? Lol. Then how big were the computers that linked them together?

blue_dragon 4y ago

> On Perseverance we use up to 300 watts to produce about 8 grams per hour of oxygen, which represents not much more than 10 percent efficiency relative to the amount of electrochemical power it actually takes to pull apart the CO2 molecule. ... It’ll still require a reliable power source, of course. According to Dr. Hecht, a nuclear reactor capable of around 10 kilowatts (currently being developed by NASA) should do the trick. To me, this is not "generating its own electricity". The device is still an incredible accomplishment, even more so because it's operating on Mars. But it doesn't generate its own electricity. They still need to connect it to a separate energy input to run the reaction. I want to clarify this just in case anybody skimming the comments gets the impression that chemically splitting the CO2 into CO and oxygen gas somehow produces enough energy to sustain the reaction without external input.

cmrdporcupine 4y ago

Multiple decades of very slow progress in fusion, space travel -- and I dunno, almost everything else -- should be enough to make it clear that Moore's Law doesn't apply to every other technological endeavour. There's no reason to assume exponential type improvements. That's not how the world works.

psadri 4y ago
First we build AGI, then we unleash it on all these other problems.
JumpCrisscross 4y ago
> no reason to assume exponential type improvements Did we read the same article? This was a test. The full-scale version “would produce oxygen at a rate equivalent to several hundred trees” with close to an order of magnitude more efficiency “because a bigger machine can run at lower pressures, saving energy on compression.” There are better reasons to expect exponential improvements than not.
ascar 4y ago
Transistor progress has been especially astonishing, but if you just look at the progress in basically all areas of living in the last 200, 1000, 5000 years there is absolutely no reason we won't see incredible further improvements in the next century. They don't have to be exponential doubling every 18 months. Steady incremental improvements compound just fine over decades. I just don't see any of your opening statement being even close to correct given the context. There was even significant progress in space travel and fusion in the last 50 years (e.g. first reactors being net positive after startup and reusable rockets with commercial space flights).
Retric 4y ago
The deeper point isn’t to compare mores law to all new technology but rather to consider how long it takes to get close to some fundamental limit. Early engines had horrible efficiency but that rapidly increased over time up to a reasonable fraction of hypothetical limits. You see this all over the place, modern guns aren’t multiple orders of magnitude improvements over guns built 100 years ago. But they are orders of magnitude better than the absolute earliest guns ever built. So, when looking at a prototype the question isn’t can it be improved but by how much can it be improved.