Climate-related disasters happening more often is a reasonable thing to worry about. I’m strongly in favor of more and better disaster preparation. However, predicting the future is very hard. For...
Climate-related disasters happening more often is a reasonable thing to worry about. I’m strongly in favor of more and better disaster preparation.
However, predicting the future is very hard. For example, who predicted the pandemic and its effects on the economy? The author of this post is wildly overconfident about their predictive ability. I’d have more respect for them if they hedged and talked about the importance of preparing for the possibility of things getting worse.
Supply gluts are also possible, because there is such a long lag between more demand and more supply, and investors often overshoot. Also people do sometimes learn from mistakes, so maybe some supply chains will become more resilient?
A terminology nit: “exponential” doesn’t mean “very fast growth.” If inflation is 2% every year then that’s exponential but not very alarming. If it were 10% every year then that’s another story.
Economists tend to assume exponential growth indefinitely, so it’s nothing new, though it is a problem from a sustainability perspective. Many apparently-exponential curves turn out to be S-curves, but predicting peaks is hard.
I’m not sure where it fits with your overall point, but many people predicted the pandemic. Not necessarily this pandemic, or the precise timing, but public health folks knew it was coming sooner...
I’m not sure where it fits with your overall point, but many people predicted the pandemic. Not necessarily this pandemic, or the precise timing, but public health folks knew it was coming sooner or later.
I agree that we can prepare for pandemics and we definitely should, and that it doesn’t require exact timing. But part of the of the problem is that, because we don’t know the timing, it can be...
I agree that we can prepare for pandemics and we definitely should, and that it doesn’t require exact timing. But part of the of the problem is that, because we don’t know the timing, it can be many decades before the investment pays off. There is a similar problem for earthquakes. We’ve been waiting for “the big one” in California for a long time. Until it happens, disaster preparation looks like waste and inefficiency. Sometimes even the people preparing for it only sort of believe they will live to see it.
For predicting what’s going to happen with the economy in the next decade, you need good information on timing, and we don’t have it. We don’t whether there will be shortages or surpluses or what prices will do. To be able to predict that you not only need the timing of upcoming pandemics, but also the timing of upcoming wars, which depends on who wins upcoming elections, and we don’t even know the candidates. It depends on how long dictators remain in power and who replaces them. It depends on whether there will be good harvests or bad. And it depends on terrorist activity - will there ever be another 9/11?
Obviously, there is no way to know these things. Anyone who tries to tell us they know the future, without hedging, is just pretending.
What do you think about this article? Is it correct? Does it miss things? What's your take on the current situation?
What do you think about this article? Is it correct? Does it miss things? What's your take on the current situation?
Prices are going to rise, probably exponentially, over the course of the next few decades. The reason for that’s simple: everything, more or less, has been artificially cheap. The costs of everything from carbon to fascism to ecological collapse to social fracture haven’t been factored in — ever, from the beginning of the industrial age. But that age is now coming to a sudden, climactic, explosive end. The problem is that, well, we’re standing in the way.
The economy is undergoing a profound shock. Unfortunately for us, it’s going to be one of the largest shocks in economic history. It’s a “supply shock,” as economists formally call it — perhaps the greatest of all time. No, I’m not exaggerating. The world can’t get microchips right about now.
The “chip shortage” is something that the world doesn’t really grasp yet, in its full importance and magnitude. It is the first climate catastrophe related shortage to hit us at a civilizational, global level.
First off, I gotta say that the language of these short examples seems greatly exaggerated and alarmist. My understanding is that the "chip shortage" is not a matter of climate change. It's a...
First off, I gotta say that the language of these short examples seems greatly exaggerated and alarmist.
My understanding is that the "chip shortage" is not a matter of climate change. It's a matter of the outsourcing the United States has done, and COVID basically halting all of that. A lot of the supply shocks that are currently happening is because of a lack of domestic production.
I've also read that the supply shock (and having a degree in Economics it's obvious to me the author is using this term to "shock" people) will eventually mellow out.
Honestly, I think that the biggest problem with the technology industry today is that - especially when it comes to hardware - practically everything is trade secret. The reason why there isn't...
Honestly, I think that the biggest problem with the technology industry today is that - especially when it comes to hardware - practically everything is trade secret. The reason why there isn't any competition for these complex process chips is not because the current players are that much better at business than everyone else, it's simply that nobody else has any clue how they're managing to do what they are doing. Look at Intel and how long they have taken to get their 7nm process up and running, and they have the advantage of the best engineers on the planet.
Is industrial secrecy worse than usual at fabs? I don’t think they’ve ever been very open. One reason there is less competition is that fabs have become increasingly expensive. At the cutting edge...
Is industrial secrecy worse than usual at fabs? I don’t think they’ve ever been very open.
One reason there is less competition is that fabs have become increasingly expensive. At the cutting edge they are doing near-impossible things. Intel made some bad bets, and it takes time to catch up.
There is literally no way to know if there are more secrets than before. That’s why they're secrets. Of course producing chips is expensive. Basic economics states that when there is less...
There is literally no way to know if there are more secrets than before. That’s why they're secrets.
Of course producing chips is expensive. Basic economics states that when there is less competition then prices will go up. Don’t you think that it’s problematic that there are so many chip manufacturers in the world but there are only a few that can be considered to be competitive?
Well, both are reasons for prices to rise. The fabs get more expensive with each generation so there is less competition and prices can go up. But there are a lot of applications that don’t need...
Well, both are reasons for prices to rise. The fabs get more expensive with each generation so there is less competition and prices can go up.
But there are a lot of applications that don’t need the latest chips, and it might be better to standardize on older tech. Nobody cares about processor speed for the chips in your car. Arguably we should stop caring about faster phones.
Why do you think we should limit innovation and accept inferior products as a 'standard'? New manufacturing processes aren't making things faster; chips are becoming faster because new...
Why do you think we should limit innovation and accept inferior products as a 'standard'? New manufacturing processes aren't making things faster; chips are becoming faster because new manufacturing processes make transistors smaller and cheaper and we are able to throw "more computer" at our computing tasks for significantly less money and power draw. When I was a kid, electronics almost universally required 5v. When the homebrew electronics craze got big, chips had just transitioned to 3.3v. Right now things are transitioning to 1.1v. Imagine if your phone ate up batteries like a Game Gear!
I get what you mean about 'faster phones', but keep in mind the reason why the world is going through phones so fast right now is not because they're getting significantly faster but because of the limited amount of support they get. And one of the largest roadblock to support is that the SoC manufacturers and designers have tons of black boxes - trade secrets - in their chips. Let me remind you that the overwhelming success of the PC platform was not necessarily because of the name recognition of IBM, but because it was a completely open system that anyone could participate in.
I don’t know why you’re bringing up 20 year old technology like the GameGear? I’m not talking about going backwards. Today’s phones have good battery life. Also, I agree that PC standardization is...
I don’t know why you’re bringing up 20 year old technology like the GameGear? I’m not talking about going backwards. Today’s phones have good battery life.
Also, I agree that PC standardization is good and it would be nice if there were similar standardization with the components that go into ARM boards. This is the kind of standardization I would like to see.
However, board-level standardization of electronic components doesn’t seem very relevant to competition between companies that run fabs? Running a factory doesn’t seem to have much to do with the open standards we are familiar with. For example, at the leading edge, apparently ASML is the sole supplier of a $120 million dollar tool needed for extreme ultraviolet lithography and only about 50 of these machines exist. Intel can buy them, but I doubt many companies can or would.
There are already a lot of embedded applications that use older-generation technologies for reduced cost, because they don’t benefit from faster chips. I think that’s mostly true of auto manufacturers, for example. Nobody buys a car based on the speed of the chips embedded in it, let alone a fridge or microwave. This isn’t inferior technology, it’s appropriate technology.
I bring up 20 year old technology because it shows how far we have come and that we have further to go. Once again, I bring up my central question: why should we limit the advancement of...
I bring up 20 year old technology because it shows how far we have come and that we have further to go. Once again, I bring up my central question: why should we limit the advancement of technology?
Are you at all interested in understanding my position? Because at this point it just feels like I'm explaining the same concepts over and over again. The concepts I'm talking about aren't just affecting chip fabs, but all the players involved in developing bleeding-edge electronic hardware.
Okay, I understand that you're making an analogy from past computing progress to future computing progress. But when we make analogies like that it's only suggestive. Maybe future progress in...
Okay, I understand that you're making an analogy from past computing progress to future computing progress. But when we make analogies like that it's only suggestive. Maybe future progress in computing won't be as much like the past as you think?
I'm not saying we should limit technology, but rather that for many applications, we wouldn't be giving up much if it turned out that we've reached a limit and the best fabs don't get any better.
Honestly, no. The demand for chips has been steadily increasing each year, and the production companies factored that in. The pandemic hit, and demand sharply increased (apparently Nvidia prepared...
The “chip shortage” is something that the world doesn’t really grasp yet, in its full importance and magnitude. It is the first climate catastrophe related shortage to hit us at a civilizational, global level.
Honestly, no.
The demand for chips has been steadily increasing each year, and the production companies factored that in. The pandemic hit, and demand sharply increased (apparently Nvidia prepared for a 10x increase with 30-series cards, and they were still massively overwhelmed) while production sharply decreased (Chip foundries had to shut down during the core of the pandemic). Together with the fact that most of the world is running on just-in-time-manufacturing which means that no one has any stocks anymore because why keep stocks when the container ship from Asia arrives every 2 weeks, means that the delivery chain is extremly fragile and is impacted greatly by such global events.
On the consumer side, scalping has become wildly rampant, which means that third party resellers are buying up stocks to resell them at an extreme markup, and no one cares because why would anyone care, the companies are selling their shit just fine.
The fact that this is written to induce fear in the user doesn't help at all.
Obviously there is context missing from the article, because I didn't quote everything from it.
Obviously there is context missing from the article, because I didn't quote everything from it.
Let’s think about that microchip shortage. What’s it really about? Well, there are three factories in which the majority of the world’s chips are made. Three factories — each hit in a different way. The one in Japan caught fire due to an equipment malfunction — apparently the blaze took hours to put out because of the conditions. The one in Texas was hit by an historic snowstorm, which knocked out power for days. The one in Taiwan is being affected by the worst drought in half a century — and microchips require huge amounts of water to manufacture.
These are all effects of climate change. They might not be the kinds of monocausal direct effects climate change deniers and American pundits look for — the hand of God roasting a factory alive — but they are very much caused by living on a rapidly heating planet. It should be eminently clear to see that when factories are freezing and burning, that is what climate change does to an economy before your very eyes. (And even if you think the Japan fire had little to do with global warming, the face of the matter is that without climate change, two of the world’s largest chip factories would still be open.)
Climate-related disasters happening more often is a reasonable thing to worry about. I’m strongly in favor of more and better disaster preparation.
However, predicting the future is very hard. For example, who predicted the pandemic and its effects on the economy? The author of this post is wildly overconfident about their predictive ability. I’d have more respect for them if they hedged and talked about the importance of preparing for the possibility of things getting worse.
Supply gluts are also possible, because there is such a long lag between more demand and more supply, and investors often overshoot. Also people do sometimes learn from mistakes, so maybe some supply chains will become more resilient?
A terminology nit: “exponential” doesn’t mean “very fast growth.” If inflation is 2% every year then that’s exponential but not very alarming. If it were 10% every year then that’s another story.
Economists tend to assume exponential growth indefinitely, so it’s nothing new, though it is a problem from a sustainability perspective. Many apparently-exponential curves turn out to be S-curves, but predicting peaks is hard.
I’m not sure where it fits with your overall point, but many people predicted the pandemic. Not necessarily this pandemic, or the precise timing, but public health folks knew it was coming sooner or later.
I agree that we can prepare for pandemics and we definitely should, and that it doesn’t require exact timing. But part of the of the problem is that, because we don’t know the timing, it can be many decades before the investment pays off. There is a similar problem for earthquakes. We’ve been waiting for “the big one” in California for a long time. Until it happens, disaster preparation looks like waste and inefficiency. Sometimes even the people preparing for it only sort of believe they will live to see it.
For predicting what’s going to happen with the economy in the next decade, you need good information on timing, and we don’t have it. We don’t whether there will be shortages or surpluses or what prices will do. To be able to predict that you not only need the timing of upcoming pandemics, but also the timing of upcoming wars, which depends on who wins upcoming elections, and we don’t even know the candidates. It depends on how long dictators remain in power and who replaces them. It depends on whether there will be good harvests or bad. And it depends on terrorist activity - will there ever be another 9/11?
Obviously, there is no way to know these things. Anyone who tries to tell us they know the future, without hedging, is just pretending.
What do you think about this article? Is it correct? Does it miss things? What's your take on the current situation?
First off, I gotta say that the language of these short examples seems greatly exaggerated and alarmist.
My understanding is that the "chip shortage" is not a matter of climate change. It's a matter of the outsourcing the United States has done, and COVID basically halting all of that. A lot of the supply shocks that are currently happening is because of a lack of domestic production.
Here's an article from the American Prospect.
I've also read that the supply shock (and having a degree in Economics it's obvious to me the author is using this term to "shock" people) will eventually mellow out.
Honestly, I think that the biggest problem with the technology industry today is that - especially when it comes to hardware - practically everything is trade secret. The reason why there isn't any competition for these complex process chips is not because the current players are that much better at business than everyone else, it's simply that nobody else has any clue how they're managing to do what they are doing. Look at Intel and how long they have taken to get their 7nm process up and running, and they have the advantage of the best engineers on the planet.
Is industrial secrecy worse than usual at fabs? I don’t think they’ve ever been very open.
One reason there is less competition is that fabs have become increasingly expensive. At the cutting edge they are doing near-impossible things. Intel made some bad bets, and it takes time to catch up.
There is literally no way to know if there are more secrets than before. That’s why they're secrets.
Of course producing chips is expensive. Basic economics states that when there is less competition then prices will go up. Don’t you think that it’s problematic that there are so many chip manufacturers in the world but there are only a few that can be considered to be competitive?
Well, both are reasons for prices to rise. The fabs get more expensive with each generation so there is less competition and prices can go up.
But there are a lot of applications that don’t need the latest chips, and it might be better to standardize on older tech. Nobody cares about processor speed for the chips in your car. Arguably we should stop caring about faster phones.
Why do you think we should limit innovation and accept inferior products as a 'standard'? New manufacturing processes aren't making things faster; chips are becoming faster because new manufacturing processes make transistors smaller and cheaper and we are able to throw "more computer" at our computing tasks for significantly less money and power draw. When I was a kid, electronics almost universally required 5v. When the homebrew electronics craze got big, chips had just transitioned to 3.3v. Right now things are transitioning to 1.1v. Imagine if your phone ate up batteries like a Game Gear!
I get what you mean about 'faster phones', but keep in mind the reason why the world is going through phones so fast right now is not because they're getting significantly faster but because of the limited amount of support they get. And one of the largest roadblock to support is that the SoC manufacturers and designers have tons of black boxes - trade secrets - in their chips. Let me remind you that the overwhelming success of the PC platform was not necessarily because of the name recognition of IBM, but because it was a completely open system that anyone could participate in.
I don’t know why you’re bringing up 20 year old technology like the GameGear? I’m not talking about going backwards. Today’s phones have good battery life.
Also, I agree that PC standardization is good and it would be nice if there were similar standardization with the components that go into ARM boards. This is the kind of standardization I would like to see.
However, board-level standardization of electronic components doesn’t seem very relevant to competition between companies that run fabs? Running a factory doesn’t seem to have much to do with the open standards we are familiar with. For example, at the leading edge, apparently ASML is the sole supplier of a $120 million dollar tool needed for extreme ultraviolet lithography and only about 50 of these machines exist. Intel can buy them, but I doubt many companies can or would.
There are already a lot of embedded applications that use older-generation technologies for reduced cost, because they don’t benefit from faster chips. I think that’s mostly true of auto manufacturers, for example. Nobody buys a car based on the speed of the chips embedded in it, let alone a fridge or microwave. This isn’t inferior technology, it’s appropriate technology.
I bring up 20 year old technology because it shows how far we have come and that we have further to go. Once again, I bring up my central question: why should we limit the advancement of technology?
Are you at all interested in understanding my position? Because at this point it just feels like I'm explaining the same concepts over and over again. The concepts I'm talking about aren't just affecting chip fabs, but all the players involved in developing bleeding-edge electronic hardware.
Okay, I understand that you're making an analogy from past computing progress to future computing progress. But when we make analogies like that it's only suggestive. Maybe future progress in computing won't be as much like the past as you think?
I'm not saying we should limit technology, but rather that for many applications, we wouldn't be giving up much if it turned out that we've reached a limit and the best fabs don't get any better.
Honestly, no.
The demand for chips has been steadily increasing each year, and the production companies factored that in. The pandemic hit, and demand sharply increased (apparently Nvidia prepared for a 10x increase with 30-series cards, and they were still massively overwhelmed) while production sharply decreased (Chip foundries had to shut down during the core of the pandemic). Together with the fact that most of the world is running on just-in-time-manufacturing which means that no one has any stocks anymore because why keep stocks when the container ship from Asia arrives every 2 weeks, means that the delivery chain is extremly fragile and is impacted greatly by such global events.
On the consumer side, scalping has become wildly rampant, which means that third party resellers are buying up stocks to resell them at an extreme markup, and no one cares because why would anyone care, the companies are selling their shit just fine.
The fact that this is written to induce fear in the user doesn't help at all.
Obviously there is context missing from the article, because I didn't quote everything from it.