All of this reminds me of the classic Venkat essay “The Zeroth Mile,” wherein a category encompassing “trash” and “noise” and “runoff” denotes “all those things which aren’t worth registering in our collective aperture of attention.”
Until they are.
What has been wild to me about the ascending relevance of rare-earths, gases, chip manufacturing equipment, etc., is how fundamentally *commoditized* they seemed to be until competitively recently. Working at NVidia, much less having the foresight to found a company like Durin, just did not register on the Trendy Societal Relevance Hotness Scale until very recently.
Which in turn makes we wonder how many of tomorrow’s most rare limiting opportunities are right now sitting beneath our collective noses, perhaps literally in the dustbin.
This is a very well-told story about a highly technical area of industrial chemistry that was largely invisible until it became vulnerable.
One question I came away with is whether there is already a meaningful response to these vulnerabilities, or whether we are still in a fairly precarious geopolitical position.
The AI factory is not just creating demand for chips. It is forcing nations to rediscover entire industrial ecosystems that had become almost invisible: gases, purification, cooling, glass, connectors, chemicals, and advanced manufacturing processes. These upstream capabilities increasingly look less like commodities and more like strategic infrastructure.
Are we protecting that strategic infrastructure yet? I’d love to learn more.
Superb and intimidating. This stuff is HARD to do. China only country with a strategic supply chain internal to the company . Still need to steal a new UViolet machine from the Netherlands
Your podcast on this topic is an absolutely amazing listen. I had no idea about the role of gas in the creation of these chips. Also, Taiwan’s 100% reliance upon Chinese gas is actually insane to hear. Lastly, Carl mentioned that the supply chain quite literally a 0 sum game. If you cannot import the gas required to make the chips, your ‘reserves’ are extremely limited because some of those gases will start to literally eat their containers.
This was a great podcast. Super informative about a topic that is mostly invisible (pun intended). One aspect that wasn't covered is how China got the manufacturing capability to begin with? Presumably much (some, all?) of the capability originally came from the outside. If so, why can't the outside rebuild it again? Commercial economics notwithstanding, surely the capabilities still exist in the US, Europe, Japan, elsewhere, or have the capabilities timed out, i.e. the talent that once knew how to do it is long retired?
There's an older version of this story: alum. In 1462, the Papal States found huge alunite deposits at Tolfa — suddenly Rome held the only major "friendly" supply of the era's essential textile mordant, until then imported almost entirely from Ottoman sources via Venice. Within a couple of years, Rome backed the mines with the threat of excommunication for anyone still buying Turkish alum. It didn't fully work — smuggling continued, and Henry VII later made a tidy profit laundering Turkish alum through England — but the shape is the same as fluorspar/NF3 today: an unglamorous chokepoint input, invisible until someone controls it, doing more geopolitical work than the finished good everyone's watching. Great piece — this kind of tracing is rare.
No mention of US efforts to fix this bottleneck? Lutnick's new Industrial Policy to find and fix all the weaknesses in our Supply Chains. Pax Silica. Deal Team Six. Chips Act is very old news at this point.
All of this reminds me of the classic Venkat essay “The Zeroth Mile,” wherein a category encompassing “trash” and “noise” and “runoff” denotes “all those things which aren’t worth registering in our collective aperture of attention.”
Until they are.
What has been wild to me about the ascending relevance of rare-earths, gases, chip manufacturing equipment, etc., is how fundamentally *commoditized* they seemed to be until competitively recently. Working at NVidia, much less having the foresight to found a company like Durin, just did not register on the Trendy Societal Relevance Hotness Scale until very recently.
Which in turn makes we wonder how many of tomorrow’s most rare limiting opportunities are right now sitting beneath our collective noses, perhaps literally in the dustbin.
This is a very well-told story about a highly technical area of industrial chemistry that was largely invisible until it became vulnerable.
One question I came away with is whether there is already a meaningful response to these vulnerabilities, or whether we are still in a fairly precarious geopolitical position.
The AI factory is not just creating demand for chips. It is forcing nations to rediscover entire industrial ecosystems that had become almost invisible: gases, purification, cooling, glass, connectors, chemicals, and advanced manufacturing processes. These upstream capabilities increasingly look less like commodities and more like strategic infrastructure.
Are we protecting that strategic infrastructure yet? I’d love to learn more.
Super insightful read!
Superb and intimidating. This stuff is HARD to do. China only country with a strategic supply chain internal to the company . Still need to steal a new UViolet machine from the Netherlands
Your podcast on this topic is an absolutely amazing listen. I had no idea about the role of gas in the creation of these chips. Also, Taiwan’s 100% reliance upon Chinese gas is actually insane to hear. Lastly, Carl mentioned that the supply chain quite literally a 0 sum game. If you cannot import the gas required to make the chips, your ‘reserves’ are extremely limited because some of those gases will start to literally eat their containers.
This podcast was fantastic! I love these deep dives by hosts excited about their work.
I work in the environmental remediation industry but on other kinds of projects and I learned a lot about this field in a short time.
This was a great podcast. Super informative about a topic that is mostly invisible (pun intended). One aspect that wasn't covered is how China got the manufacturing capability to begin with? Presumably much (some, all?) of the capability originally came from the outside. If so, why can't the outside rebuild it again? Commercial economics notwithstanding, surely the capabilities still exist in the US, Europe, Japan, elsewhere, or have the capabilities timed out, i.e. the talent that once knew how to do it is long retired?
Great timing for this post, as China just announced new export controls on helium. Now I understand why.
There's an older version of this story: alum. In 1462, the Papal States found huge alunite deposits at Tolfa — suddenly Rome held the only major "friendly" supply of the era's essential textile mordant, until then imported almost entirely from Ottoman sources via Venice. Within a couple of years, Rome backed the mines with the threat of excommunication for anyone still buying Turkish alum. It didn't fully work — smuggling continued, and Henry VII later made a tidy profit laundering Turkish alum through England — but the shape is the same as fluorspar/NF3 today: an unglamorous chokepoint input, invisible until someone controls it, doing more geopolitical work than the finished good everyone's watching. Great piece — this kind of tracing is rare.
No mention of US efforts to fix this bottleneck? Lutnick's new Industrial Policy to find and fix all the weaknesses in our Supply Chains. Pax Silica. Deal Team Six. Chips Act is very old news at this point.