IP25: Moving Fast And Making Things. How AI Will Drive A Manufacturing Revival

And Why Tech Should Talk More About Its Role In Boosting Manufacturing

Apologies for the hiatus in posting. My spare time has been taken up with a large scale and short-notice book writing project. More on that soon…

Remember if you like this post, please share with others who might be interested. And please use the like ♡ button if you can - seemingly Substack’s algorithms love the like button!

Moving Fast And Making Things. How AI Will Drive Manufacturing Revival

This week I’m keen to look more deeply into the ability of tech, robotics and AI to transform manufacturing. And props to the always excellent Oren Cass of American Compass, who used the phrase “Move Fast and Make Things” in a recent piece with Dean Ball on American manufacturing. I couldn’t resist using that superb title for this piece as well. 

In short, my argument is that manufacturing (notably around semiconductors, chips and batteries) is key to tech, but also that advances in AI are integral to attempts to reindustrialise and bring about a revival in manufacturing. The conclusion for people who work in tech is that the centrality of AI to a manufacturing renaissance (an idea that has cross political support) should be a much more important part of messaging and narrative.

Tech is fundamental to a manufacturing revival

There is a growing societal consensus of the importance of boosting manufacturing. This is, correctly, seen as being important for productivity, economic resilience and national security. It’s also essential for creating good, well paid jobs and helping to revive towns and cities that have been struggling since the decline of heavy industry.

What isn’t considered enough is how important tech is to these attempts at manufacturing revival. Most fundamentally, one of the most important catalysts for attempts to strengthen manufacturing is the need for access to semiconductors, chips and other elements that are essential to the tech and AI revolution. 

So, a resilient and reliable manufacturing is essential for continued advance in tech. But the opposite also applies - continued advances in tech are essential for a resilient and reliable manufacturing. 

AI will help to improve manufacturing as we know it today. More intriguingly and excitingly, AI will help to transform manufacturing processes through the production of new materials and the understanding of how chemicals and proteins are formed. More intriguingly still, the next generation of AI-driven robotics could transform manufacturing in an even more revolutionary way. The rest of this note will consider why manufacturing still matters, why tech and AI is essential to reviving manufacturing and why the tech community should make this a core part of its messaging.  

The Reindustrialisation Imperative

For years, it was accepted by politicians and economists in Western democracies that manufacturing industry would be a smaller part of the economic pie. An emphasis on comparative advantage in a world of economic globalisation meant that many countries started to focus on services, rather than manufacturing. The model became one of “just in time”, where global supply chains meant that products were supplied at the last minute, with several major products requiring multiple elements from across different supply chains.  Across Europe and North America, deindustrialisation became the order of the day, with particular impact on local economies in places like Ohio, the North of England, Nord-Pad-de-Calais in France and the Ruhr region in Germany.

Now the political and economic imperative has shifted from deindustrialisation to reindustrialisation. In France, President Macron argued that “the battle for reindustrialization is key on an economic level, it is key on a geopolitical level, and it is key on a political level for the unity of the nation…We must accelerate and go much further.” The Australian government set out a goal to “rebuild Australia’s industrial base”. Under President Biden and VP Harris, the United States has committed to “revitalize American manufacturing” - a priority shared by Donald Trump and JD Vance.

Partially, geopolitics has forced this change. Decoupling with China and the increased importance of access to physical goods, such as semiconductors, has elevated the priority of boosting domestic manufacturing capacity.  The war in Ukraine showed the limits of interdependence (which Russia latterly defined as “weaponised interdependence”), with the inflationary shock being felt across Europe for years. Difficulty in accessing goods, ranging from masks and vaccines to semiconductors, during Covid emphasised the importance of economic resilience in supply chains, as did the temporary blockage of the Suez Canal. Resilience and “just-in-case” became much more important than openness and “just-in-time”.

Politics has also played a role in the changing dynamics, as is illustrated by President Macron’s phrase that reviving industry is “key on a political level for the unity of the nation.” Populist movements in various countries have seen their biggest support in areas that were once reliant on heavy industry.  This dynamic, along with what Piketty described as the Brahminisation of the left, meant that the loyalties of working class voters has shifted from mainstream left to populist right. This populist surge has meant that both left and right have turned to reindustrialisation - the left to win back their former heartlands and the right to entrench their gains. An increasingly important green tinge to politics, particularly on the left, also means that moving production closer to consumption and reducing the carbon footprint might become increasingly important politically. 

Economics has also played an important role beyond the economic imperatives of geopolitics. The shift to reindustrialisation is not a nostalgic political kneejerk, but a legitimate response to economic needs. Notably, recent evidence has shown that a strong manufacturing base is important for boosting productivity and creating value from R & D. OECD data, for example, has shown manufacturing to be a greater contributor to productivity growth than services.  Manufacturing accounts for fewer than a tenth of the jobs in the UK, for example, but represents almost two thirds of its R&D investment. The only other sector that comes close to this is tech - note the historical importance of Bell Labs in the United States and the crucial R&D contribution of companies like Alphabet today.  Manufacturing also creates what Shih and Pisano describe as the “industrial commons” - hubs and clusters of innovation that drive up productivity. 

Tech and manufacturing are mutually beneficial

As noted, a strong manufacturing sector is important in providing a secure supply of semiconductors, chips and other key elements of the AI revolution. And resilient supply is even more important in an age of decoupling and geopolitical uncertainty.

What is less well understood however, is the importance of the AI revolution to a manufacturing renaissance. AI is crucial to three parts of reindustrialisation. The first is, simply, making existing processes more productive and more effective. The second is potentially revolutionising the production of materials that are the raw ingredients of manufacturing. And the third is via a dramatic advance in the use of AI-driven robotics.

I explained the benefits around AI enhancing existing processes in IP24:

AI can help manufacturers in were once tasks that constrained productivity growth, such as predicting when machines need maintenance; taking a more detailed approach to quality control; and identifying inventory blockages and supply-chain needs. It is also able to predict peaks and troughs in demand, meaning that there is no longer any need to make the false choice between efficiency and resilience.

The second potential of AI in manufacturing could be remarkably transformative. In my first book, Little Platoons, I wrote about how my home town of Consett became a global capital of the steel industry decades after a group of Prussian swordmakers settled in this part of the North of England - attracted by the iron ore and the softness of the water. As the Industrial Revolution accelerated, places with similar resources first industrialised and then deindustrialised.

AI could help in the discovery of new materials, which could transform the manufacturing of products ranging from consumer goods to defence products. Indeed, many next generation semiconductors, batteries and other devices require the development of new materials. This potential has meant that many countries have invested heavily in this materials research, with a Georgetown CSET report finding that AI research and material represented the biggest source of US government grants to private industry in the six years to 2023. 

The German headquartered Freudenberg Group, for example, is using machine learning to help speed up the process of material discovery, with many of their companies directly linked to the manufacturing process. Deepmind’s Graph Networks for Material Exploration (GNoME) project is one of the most exciting developments in the discovery of new materials. In their Nature paper explaining the project, Deepmind explain that:

Novel functional materials enable fundamental breakthroughs across technological applications… From microchips to batteries and photovoltaics, discovery of inorganic crystals has been bottlenecked by expensive trial-and-error approaches. Concurrently, deep-learning models for language, vision and biology have showcased emergent predictive capabilities with increasing data and computation… graph networks trained at scale can reach unprecedented levels of generalization, improving the efficiency of materials discovery by an order of magnitude.

The results of the GNoME tool are simply astonishing: 

“The discovery of 2.2 million new crystals – equivalent to nearly 800 years’ worth of knowledge… Of its 2.2 million predictions, 380,000 are the most stable, making them promising candidates for experimental synthesis. Among these candidates are materials that have the potential to develop future transformative technologies ranging from superconductors, powering supercomputers, and next-generation batteries to boost the efficiency of electric vehicles.”

The impact of such discoveries on advanced manufacturing could be remarkable, as could continual advances in fields such as 3D Printing. 

Robotics represent the third area where AI could help drive a manufacturing renaissance. This great, Olympic tinged example, of a Deepmind trained robot playing table tennis is a compelling example of what AI driven robotics might do. 

Whereas the most recent generations of robots used in manufacturing often have a single pre-programmed purpose. As such, some of the biggest AI investments by VCs so far this year have been in robotics, with the likes of Skild.AI and Physical Intelligence having a focus on “bringing generative AI to the real world”. Physical Intelligence plans “to create software that can add high-level intelligence to a wide variety of robots and machines.” Their CEO says that they are aiming to “bring AI to the physical world with a universal model that can power any robot or any physical device basically for any application.”

Much more powerful robots could clearly help to turbocharge a manufacturing revival. Many robotics in factories at the moment (including in early adopting countries like South Korea) only have minimal use of AI. They tend to be only programmed for a single task, cannot progressively learn and cannot speak to other robots. AI driven robots that can learn and communicate could transform the factory environment.

Powering a manufacturing revival should be a core part of the tech narrative

For many years, tech has had a perception of being “Exiled from Main Street”. A Brookings study, which proved to be highly influential on both left and right (and cited by politicians across the spectrum), found that “just five top innovation metro areas—Boston, San Francisco, San Jose, Seattle, and San Diego—accounted for more than 90% of the nation’s innovation-sector growth during the years 2005 to 2017.” 

The idea that big tech is somehow only bringing benefits to metropolitan areas is a myth that has taken hold. Overcoming this kind of myth needs a strong narrative and a compelling story. The skilling narrative detailed in IP24 is an important part of this, but the narrative of AI-driven manufacturing revival is also essential. 

The narrative of an AI-driven manufacturing renaissance is an ideal narrative to overcome this metropolitan elite perception. It can be seen as tech playing a key role in creating jobs and industries in the industrial heartland, building well-paid, secure work and strengthening communities that have felt left behind for years or even decades. 

Boosting manufacturing is popular amongst the public and transcends political divides. Tech playing its role in creating good dignified work in strong communities would be an important statement of an impact that goes well beyond the creation of consumer products and services.

Tech making a big impact by playing a key part in reviving manufacturing is a powerful message that can explode myths and go beyond metro-elite cliches.

AI needs a strong manufacturing sector. A manufacturing revival also needs advances in AI. The importance of AI to manufacturing revival is a key message for the tech sector to communicate.

Afterword. More musings on tech optimism

Thanks for all of the great feedback into the recent Substacks on techno-optimism and five reasons to be optimistic about AI, jobs and skills. I’m keen to come back to both subjects in more detail over the next few weeks, but there were a few things that I’ve listened to and read over the past few weeks that provide an interesting angle into the techno-optimism discussion.

The first was Ezra Klein, who said on his podcast that when people describe him as a tech-optimist, his response is always that he is a “political realist”, meaning that most of the priorities of progressives can only be achieved via technology. This is an interesting thought about tech-optimists moving focus and messaging from the perils of technology from its promise. This case was also made by former Prime Minister, Tony Blair, who correctly made the case that the only way for the UK to prosper is by fully embracing the promise of technology and AI. 

The second techno-optimist point was about Margaret Thatcher, a politician I’d never previously thought of as a techno-optimist. I’m reading Dominic Sandbrook’s excellent Who Dares Wins, which is the latest in his tremendous series of British post-war histories. In the book, Sandbrook talks about Thatcher’s zeal for ensuring that the UK made the most of the opportunities of the microchip, arguing that “we must be way up front in the new industries… for new technologies bring new opportunities as well.” This involved, in 1981 appointing Kenneth Baker as Minister for Information Technology, getting computers in schools and having a prime-time BBC programme based on learning computer skills, which gained over 7 million viewers. One of Baker’s first acts after being appointed IT Minister was to describe fears of the microchip leading to job losses as “absolute balls”. The poster child of this wave of British techno-optimism was Acorn computers, which is sadly now defunct but was amongst the first to make the personal computer in the home a reality and lives on, with its joint-venture with Apple that launched in 1990 now known as ARM. 

Klein and Mrs Thatcher are a useful reminder that tech optimists can come from across the political spectrum. Something that unites optimists across the spectrum today is a belief in the importance of manufacturing. And nothing will be more important than technological advance to ensuring a manufacturing resurgence.