I’ve deliberately tried not to write too much about AI, because the signal is drowned out by the noise. But I think the picture is getting clearer now. This week on The Next Wave, I’m going to republish versions of posts in my newsletter, just two: one from last summer, and one that goes live this week.
—
Just by way of a thought: if the current rise of the cluster of technologies under the label ‘AI’ is not the beginning of an entirely new technology wave, but actually the culmination of the digital wave that began in the 1970s and accelerated at the turn of the century?
I’ve been thinking about it for a while in a vague kind of way because I haven’t been able to To see the business model That supports huge investment in AI in the US. (I did Wrote about this before here.)
It’s a long way for some of the pieces from Nicholas Collin, strategy and innovation blogger, who thinks the same thing, but does it much more coherently. He called itLate cycle investment theory‘
Perez model
Like me, he’s a fan of the work of academic Carlota Perez, who built on Christopher Freeman’s work to show how technology and money interact to create new long waves of investment, starting with canals and cotton, that last 50-60 years. (He calls them ’emergences’ because each technology, unlike ‘waves’, embeds itself in society and its infrastructure.)
Two recent waves are a car/oil wave, which began in 1908, and information and communication technology, which began in 1971.
I’m not going to go into all the theory of the Perez model here – if you want to, it’s online and I have it It is written about elsewhere—but what is relevant for the current discussion is that it follows an S-curve, and the first half is slow, as new infrastructure is ‘installed’, and some of it is under the radar. The Internet was a closed academic network for most of the first part of its S-curve.
‘Deploy’ from Infrastructure
Halfway through, after a lot of infrastructure is built, and usually after a financial crash that causes some of the investors in that infrastructure to lose their shirts, ‘minus’ companies with real customers and business models take over and grow exponentially before crossing the market cap and becoming a normal business. And investors who made huge profits from that boom period began looking elsewhere—for the technologies that would fuel the next boom.
The reason I like Perez’s version is that his model has gained a lot of explanatory power over the past 25 years as I’ve watched the technology sector evolve.
That’s a long way into Colin’s argument, and Let me quote directly from his first article:
Viewed through a late-cycle lens, today’s markets show signs that we have entered the maturity phase of the computing and network revolution. The theory, therefore, leads to specific, testable predictions about where capital will go and which strategies will outperform.
Three indicators
He points to three indicators in the technology sector that support the observation that we are in a ‘late cycle’:
- The 2022 startup funding decline wasn’t just a correction — it could be structural. As investor Jerry Newman argues in his landmark Productive uncertaintyStartups rely on uncertainty as a competitive edge. When good ideas become clear to everyone – including those responsible for good funding – the startup model faces practical pressures.
- Then came AI, revealing new dynamics. ChatGPT’s breakthrough didn’t come from a garage startup but from OpenAI backed by Microsoft’s massive computing power. Google, Meta, and Amazon received billions of responses. This pattern—big technology deploying huge capital against well-understood problems—fits well with late-cycle theory.
- Most notably, platform saturation now appears almost complete. Digital transformation has reached most sectors where computing and networks can possibly work. which remains-Healthcare delivery, education, construction, Government service– May reflect the natural limits of the paradigm, not untapped markets. (his emphasis)
Optimizing existing systems
In Second article —behind something of a paywall—he specifically looks at how AI is being deployed, and I’m going to quickly quote/paraphrase the visible bits of it here.
Late-cycle investment theory Suggests AI is an advance in the computing and network era of efficiency, not the beginning of a new one. just like Lean manufacturing refined mass production in the 1970s without replacing itAI optimizes an existing paradigm rather than creating a new one
Colin has done a lot of analysis here, and he has gathered a lot of evidence which he shares. I’m not going to spend too much time on this, because I’m more interested in the big strategic questions that arise if he’s right.
What a new technology wave looks like
But it is worth summarizing some observations. First, at the beginning of the emergence of a new technology, you don’t know it’s happening. You will later understand the decisive moment, the moment an innovation transformed the cost structure (the spinning genie, Watt’s condensing engine, the Ford production line, the microprocessor). But with AI, the moment was very visible, to the point of being choreographed.
Second, the amount of capital investment is off scale. In the early stages of a growth, investment is vague and not fully understood—the sector exists but it is not yet fully legible.
And third, Colin suggests that AI allows computing to reach sectors that have in some ways resisted it:
Like lean manufacturing, which extended the dominance of mass manufacturing for decades through efficiency gains, AI does not mark the end of computing but rather its maturation. The technology spread to previously untouched sectors, creating the illusion of radical innovation while actually representing the ultimate triumph of the physical economy of computing and networks.
late installation
It’s worth pausing here. Although Perez makes possible subsequent leaps from the date of innovation of each of his surges, the older wave has a kind of ‘late deployment’ phase while the newer one is still in the early stages of development.
Late deployment: So although the ICT boom is from 1971, most of the final innovations in cars/oil growth are also from that time. In the UK at the time, there was still a massive road building program of motorways and ring-roads, and this enabled the rise of long-distance logistics, out-of-town big-box retail and the edge of city business parks. Colin’s argument that AI is the equivalent of high street and big box retail is different, but the kind of transformative change about embedding technology more deeply ultimately creates a new and distinct form of abundance.
There is also social pushback—campaigns against major ring road projects have begun in the UK In the late 1960s and early 1970s. And maybe we’re seeing something about AI. US map of local pushback against data centers from Data Center Watch covers states in red and the entire country in blue. people think hatred Google inserts AI tools into its search results, and hates that it’s impossible to stop. It does not speak to an exciting technology that is being adopted by its users. A note by Ted Gioia on his music blog said that:
Most people wouldn’t willingly pay for AI—just 8% According to a recent survey. So (tech companies) have to bundle it with some other essential product.
or as As Ed Zitron recently noted Concept:
Idea raised its business plan from $15 to $20 per user per month Thanks to its new “AI features”, which I imagine will appeal to previous business customers who don’t want “AI agents” or any of that crap but want things like single sign on and premium integration. The result? Profit margin decreased by 10%. Great job everyone!
normal return
This is important for a few reasons. First, late-stage post-deployment technologies provide returns on investment, but they are normal returns, not increasing returns.
But secondly it sheds a different light on the fact that there is currently a ‘business model war’ going on between China and the US through their different approaches to AI.
I think we know a lot about the American model. This aspect is driven by a transhumanist ideology the raptureAs OpenAI’s Sam Altman reminds people every week of the year.
The Chinese model of AI
As the Exponential View newsletter Sunday explained, Citing the policy organization RANDThe Chinese model is completely different:
In Washington, the AI policy discourse is sometimes coined the ‘race to AGI’. In contrast, in Beijing, the AI discourse is less abstract and focuses on economic and industrial applications that can support Beijing’s overall economic objectives.
Azim Azhar of EV adds some gloss:
Chinese teams … release lightweight open-source architectures and partner with experts in areas such as healthcare analytics (Edu Tech) and adaptive learning (Squirrel AI)
This is partly driven by constraints: China has far less computing power than the US and has to build leeches. This means its model is much more exportable. But the important point here is that if AI is a late-stage technology and not the next big wave of innovation, the Chinese model fits the bill. Perhaps we shouldn’t be surprised: unlike most countries, a third of the full members of China’s Central Committee are technocrats.
–
Read now:
–
This is a slightly updated version of this article that was first published on me Just Two Things Newsletter.
–
Previously published with thenextwavefutures.wordpress.com Creative Commons License
***
If you believe in the work we’re doing at The Good Men Project, please join us as a Premium Member today.
All premium members can watch The Good Men Project without any ads.
Need more information? A full list of benefits is here.
–
Photo by Igor Omilaev on Unsplash
