Hacker News Reader

Comment by mhartl | Hacker News Reader

Before people freak out about exponential growth, please consider John McCarthy's Slogan. (Yes, that John McCarthy; Lisp isn't the only thing he's done.)

  He who refuses to do arithmetic is doomed to talk nonsense.

http://www-formal.stanford.edu/jmc/progress/index.html

Based on reactions to this video (among other things), I'd add Hartl's Counterpoint:

  Just because you're willing to do arithmetic doesn't 
  mean you're not talking nonsense.

In case you're wondering why you shouldn't freak out, I offer you this: http://en.wikipedia.org/wiki/Logistic_function

View on HN

Huh. This guy is simply making the very basic point that politicians, economists and bankers selling us "growth" as a solution to unemployment, debt, etc are plainly lying. Growth (population, GDP, energy availability, etc) will stop at some point. An economy that relies entirely on growth to merely function is doomed to fail spectacularly at some point in time.

I agree that population growth cannot go on forever (although I consider doubling the current number quite likely). However I see no reason why energy availability can't keep growing for many many centuries to come. We aren't tapping the full potential of nuclear fission, to say nothing about fusion. Things like solar energy in all its variations is also a source we basically aren't using at all. Give us 50-100 years an we'll no doubt come up with some new source of energy we haven't even considered yet.

> However I see no reason why energy availability can't keep growing for many many centuries to come. We aren't tapping the full potential of nuclear fission, to say nothing about fusion.

Nuclear fission is currently on decline. We aren't even building enough power plants to maintain its output, nor are we likely to do so in the next 20 to 30 years. Industrial nuclear fusion is decades away, probably a century.

From the 18th century onwards, we went from tapping ever better energy sources (denser, lighter): animal power, coal, oil. There aren't any better energy source than oil available now, or in the foreseeable future. At some point, we'll have to do with /less/.

"At some point, we'll have to do with /less/."

This part of the long business cycle always produces a lot of arguments for austerity. So far they have always turned out to be spectacularly wrong.

"Nuclear fission is currently on decline."

Nuclear power, both fission and fusion, is currently in a renaissance and we appear to be laying the groundwork for the next stage of advances: gas- and metal-cooled reactors, superconductors, mechanical compression systems, reliable mode-locked lasers, dielectric wall and wake field particle accelerators, electronics that can do real-time gigahertz feedback, etc.

When the next long expansion gets into full swing, I predict that the nuclear logjam will break loose.

"There aren't any better energy source than oil available now, or in the foreseeable future."

There's a lot of solar power in space. A lot. And literal mountains of metal, richer than the best terrestrial ores and pre-smelted too. At some point technology will tip the profit margin positive on exploiting it.

> So far they have always turned out to be spectacularly wrong.

So far, there weren't 7 billions humans neither.

> Nuclear power, both fission and fusion, is currently in a renaissance

Absolutely not. This is GE and AREVA PR, but doesn't reflect the truth. We aren't building enough nuclear plants now to compensate for the future closing of the old ones. The numbers aren't even ramping up fast enough. As time goes by, there will be less and less active nuclear plants.

Fusion research is blooming, but it didn't make much progress, particularly compared to the huge effort (see ITER).

> There's a lot of solar power in space. A lot. And literal mountains of metal, richer than the best terrestrial ores and pre-smelted too. At some point technology will tip the profit margin positive on exploiting it.

This is science-fiction. I'm pretty close to say this is complete BS. "Space ore" simply can't happen unless you find the way to harvest the power from some magic ponies yet to discover.

I totally agree that we aren't building enough nuclear power plants, but that is purely for political reasons. The research is still happening and once there is a change in the political climate we'll be able to ramp up nuclear power over a few decades.

Also if we look back at the past 1000 (or even just 100) years of human history, we've been pretty good at discovering and harvesting 'magic ponies' of all shapes and sizes.

> we'll be able to ramp up nuclear power over a few decades.

Maybe if thorium comes on the radar.

> we've been pretty good at discovering and harvesting 'magic ponies' of all shapes and sizes.

Well, not that much. By the end of the roman republic, most of what made the industrial revolution possible was available, but it didn't happen until 18 centuries later.

Except about 200 years of science and math starting from the end of the Renaissance Period.

The Antikythera mechanism (2 century BC) basically reset what we believed of science, mechanics and mathematical knowledge during antiquity. That, and some other things : Hero's Eolipyle, Nero's emerald monocle, Provence "industrial" mills... They had the ability, they lacked the will.

They lacked metallurgy, banking, agricultural technology... The Industrial Revolution was built on hundreds of years of slow progress and access to tons of natural resources. The Aeliopile was a toy that opened doors; it wasn't going to power any steamships.

There's a strange parallel to the evolution of life on Earth. Took about a billion years before the simplest single-celled organism developed all the basic mechanisms/components, and then a few hundred million years from complex multicelled organisms, and finally a huge explosion in diversity in the last hundred million years. I wonder if evolution and technological progress plotted on the same relative logarithmic scale would overlay each other?

The ancient Greeks even had an initial stream engine, but never put it to practical use, relegating it to parlor/temple game.

http://en.wikipedia.org/wiki/Aeolipile

That's what I mean : they had the ability to go to industrial age, they lacked the will, the impulse.

Also, they had slavery.

While Thorium reactors are cool and all, they aren't necessary. It's not like we're about to run out of uranium any time soon. Hell, you can take your basic 1970's uranium reactor, incorporate everything we've learned about nuclear safety over the past 40 years, throw in some generic improvements to basic turbine technology and have a fully functional power plant that will do a splendid job for decades to come.

And still no proper way to manage nuclear waste.

Here are some options. Forgive my brevity, but I'm tired of people simply asserting that there's no way of managing nuclear waste.

* Throw it away. Deep-sea subduction zones work nicely for this. Tremendously wasteful, though; that "waste" is valuable fuel.

* Recycle the waste to decrease its volume. This isn't a full solution, but it makes the waste cheaper to store for a while.

* Store the waste for a while, and later we can use it in breeder reactors or fusion/fission hybrids. There's so much useful U-238 in that slightly-used fuel. After this, the waste is pretty much dead.

I'm tired of this toon and particularly of the general political atmosphere on HN (technology will solve all problems! yay!).

If it's so easy to get rid of nuclear waste, why didn't we do it yet?

GP just said. Because getting rid of it would be stupid - the current approach is building long-term storage.

So far, technology has solved all problems. What else is there?

I know! We can vaporize it and spew it into the air, like the coal industry already does with their radioactive waste!

"We aren't building enough nuclear plants now to compensate for the future closing of the old ones."

Which is an intentional choice, arising from a combination of ennui, defeatism, and witch hunting. Those political movements seem to have run their course.

"Fusion research is blooming, but it didn't make much progress, particularly compared to the huge effort (see ITER)."

Most of the tokamak megaprojects were done to give academics a safe way to demonstrate activity. ITER is the plasma physics version of string theory. In my opinion, the real progress has come from small quiet projects involving vacuum physics and power electronics.

"'Space ore' simply can't happen unless you find the way to harvest the power from some magic ponies yet to discover."

It's metal, not ore. If you can lay your hands on asteroidal metal, you can turn it into useful articles with a blacksmith's forge. Fancy alloys and precious metals are more effort, but often easier than dealing with crappy terrestrial ores.

The magic pony is to cancel the failed Shuttle, yet another stagnant megaproject. And avoid replacing it with other doomed megaprojects. The barrier to space activity is heavy lift capability to get there at all. Once you can get above the atmosphere cheaply enough, the solar system is your oyster. (It's like Fed Ex. The capital investment to build Fed Ex is ludicrous, but once somebody has the balls to do it, entire industries rise up from nowhere to exploit it.)

> The barrier to space activity is heavy lift capability to get there at all.

Pipe dreaming. The barrier to space is the _tremendous_ amount of energy required to get anything to low-earth orbit, let alone getting raw material going back from Jupiter's vicinity. Even with 1000 times better technology (which won't arrive soon), your asteroid iron still will be costlier than platinum.

Science-fiction is mostly fiction, you know.

Nope. The energy cost is trivial. The cost of joules for a vacation to Europa is within upper middle class aspirations. It only costs so much because our lifters are hideously inefficient.

The reason rockets are so inefficient is that we don't know how to keep cracks from spreading through our structural materials, so we have to derate their strength by huge factors to get reliability. Nano- and micro-composite materials look like they will be able to vastly improve this situation (and indeed already are in the latest jetliners), and the rocket equation gives big big cost savings for weight savings. High strength materials are also good for rotovators and other semi-passive lifters.

Another big opportunity is to shift some of the propulsion to the ground. It should be straightforward to shine launch lasers up the tail of a standard rocket and reduce its propellant requirement by many percent, which the rocket equation then multiplies into a large cost savings. Exclusively laser launched vehicles are also possible but speculative at this point (they need visible light lasers to get range but those are still too inefficient). Most of the necessary technology exists or can easily be built. All we lack is the optimism to make the investment.

As for getting metal back, the asteroid belt is energetically close despite the distance. A small nuclear reactor and patience will do the job.

> Nuclear fission is currently on decline. We aren't even building enough power plants to maintain its output, nor are we likely to do so in the next 20 to 30 years.

That's a choice, one that we could change.

There are no technical or economic obstacles to building nuclear plants for far less money and in far less time that are just as safe (or safer) than what we have now. (France is an existence proof for much of the potential savings.)

There are, however, legal and political obstacles.

> Nuclear fission is currently on decline. We aren't even building enough power plants to maintain its output, nor are we likely to do so in the next 20 to 30 years.

Define "we". Take a look at China's nuclear power program sometime; they've brought the cost of their CPR1000 plants down to where they're cheaper to build than coal plants of the same size, and they're well on their way to do the same for the more modern AP1000 reactor design. They've got a bunch of new nuclear plants under construction, with more on the way as they pick up momentum. We're talking about 10-15 large reactors coming on line every year, with the production rate increasing as they ramp up.

Also, they've got their first commercial-scale pebble-bed reactors being built right now. If I remember correctly, they're also looking at incorporating thorium into the fuel mix.

Also, they're working on 40% and 70% more powerful versions of the AP1000 that will not add much to the cost. Nuclear power in China is going to be seriously cheap.

Also, they're working on fast breeder reactors, which use uranium much more efficiently. They're building two BN-800 reactors, working together with Russia, since Russia has been profitably operating fast breeder reactors in the BN series for decades.

If you're only looking at the US and Europe, then sure, the Nuclear Renaissance is looking a little feeble. But the US and Europe are not the whole world.

nuclear power could deliver far more energy per year than is used yearly now for thousands of years. Innovation makes energy more useful. When the potential for growth these two cause is exhausted there's scifi stuff; solar power from satellites around the sun; fusion; etc.

Right, but it's still finite to some degree, and selling people on growth which we cannot guarantee is possible is just plain irresponsible IMO.

> Right, but it's still finite to some degree, and selling people on growth which we cannot guarantee is possible is just plain irresponsible IMO.

Do you plan as if disk drives won't get more capacity? That ram and cpus won't get cheaper?

It's only irresponsible if you promise growth that won't happen. It's irresponsible to ignore growth that will or can happen.

Is it still lying if you don't know that you are wrong?

Dr. Albert A. Bartlett is fully cognizant of that point. He says in part 8 of that video:

  "Except for the petroleum graphs, the things I tell you are
  not predictions of the future. I am only reporting facts and
  the results of some very simple arithmetic." (1:22)

He acknowledges the potential fallibility of his estimations. He goes on to say:

  "Please you check the facts. Please check my arithmetic. if
  you find errors, please let me know. But if you don't find
  errors, I hope you take this very very seriously."

The video also makes the point that people are ignorant of the exponential function's effect, believing that constant growth is possible, when it clearly is not.

The Dr. also mentions that 0% growth is inevitable, so that is covered - but I had no idea it was called the logistic function, thanks.

In the context of population growth, the logistic function means that something has made it stop. He dealt with this is the second video - he says we'll get zero population growth one way or another, because exponential growth forever is impossible.

Around minute 7 of the second video (did you actually watch it?) he points out that many of the checks on population growth are generally considered bad things - war, famine, disease (the alternative being birth control). This is hardly a reason not to freak out.

did you actually watch it?

Yes.

many of the checks on population growth are generally considered bad things - war, famine, disease

The countries with the lowest population growth (e.g., Germany, Italy, Bulgaria, Japan) are among those with the least war, famine, and disease. Something else is going on.

Do you notice that the countries with the lowest population growth were severely battered in wars during the last century?

Italian birthrates aren't low because of Il Duce. Once people reach a certain level of wealth, they voluntarily stop having lots of kids.

There were many other reasons listed, including low immigration, small families, contraception/abortion, abstinence and pollution.

This checks do apply do the countries you cited, but the list is still far from comprehensive. They are far better than war, famine and disease, but still regarded as bad things, which is the point.

The dominant source of low birthrates in rich countries is people choosing to have smaller families.

To the point: What seems like exponential growth often turn out to be better described as S-curves.

The reason that the curve levels out for biological populations is that the rates at which members start starving to death or killing each-other increases to cancel out the birth-rate. That does seem to be of some concern.

Reply on news.ycombinator.com