ENTRY
[ESC]DOC TAYROC'S [UNSOLICITED] ECONOMICS HOT THOUGHT
Marx and the Death of the Author XVIII: Capitalism in Crisis IV: The House that Watson Built
In my work as an economist I wasn't always the 'AI guy'. Especially since so much of my work predates the current AI bubble. At varying points in my work I've focused on Foreign Direct Investment (FDI), international trade, intergovernmental organisations, non-AI digital automation, and 'platform economies'. When I started my PhD, the focus was on broader digital automation and the 'gig' economy and its impacts on waged labour in Nigeria. AI was an after thought, with far more of the initial drafts of my PhD focusing on 3D printing rather than AI. There was a brief examination of AI, mostly focusing on IBM's Watson, the earliest versions of ChatGPT, and Google's Go-playing AlphaGo, which was the height of 'AI' development when I started my PhD, but the AI bubble hadn't arrived yet.
It wouldn't be until I was well into my PhD that ChatGPT 3 launched and started the current AI bubble, which necessitated slight, albeit hasty, changes to my field work methodology and saw the AI parts of my PhD grow, displacing 3D printing. However, as I reviewed the results of my field work, I began to notice a considerable pattern: managers were planning and altering business and hiring policy based on AI despite not having any clear ideas of what exactly AI could, or more importantly, couldn't do. AI's presence in the Nigerian economy had little to do with what AI itself was doing, but everything to do with the perception, the idea, the promise of what it could, maybe, someday, probably do. For the fourth time in this miniseries, an insane amount of capital was moving, and people's lives pre-emptively ruined, over speculation. No measurable concrete change, but measurable economic activity regardless. Statistics without any basis in reality, just the hopes and dreams of the capitalist class.
This time might feel a bit different on the surface, but that's only because of the additional layer of abstraction mixed with the insane amount of 'three card Monty' that the bourgeois class is currently playing. The tulip crisis had a basis in reality, no one is contesting the real nature of tulips nor that tulip bulbs do indeed eventually become tulips. The benefits of trains during rail mania was obvious, even the earliest trains could pull a remarkable amount of freight at a relatively remarkable speed, at least compared to the horse-and-buggy arrangements of old. The over production of goods in the lead up to the 1929 crash was a testament to how refined the manufacturing process had become, and whilst there may have been more Ford Model T's produced than were demanded, no one contested the material reality of a Model T. Even the 'dot com' crisis had basis in obvious material reality, computers and the internet were measurably, incontestably real, and were without a doubt changing, mostly for the better, human life.
AI is stickier, as the goal posts for what constitutes AI keep moving. You've likely noticed that the current AI boom seems to have more in common with the crazes around radium and electricity at the turn of the last century, which saw no shortage of grifts around the lack of understanding the average person had around these miracle discoveries. These grifts lacked immense impact on the economy since you seldom had entire industries rebuilding themselves around fundamental misunderstandings of radium and electricity. Both radium and electricity are real, but neither have half the medicinal or capitalism changing implications that were claimed at the time. AI, on the other hand, has seen a sizeable portion of the economy centre around it, not unlike subprime mortgages and the 'cheap credit' of the 1980s-2000s. Most of the capital centring around AI has little to do with the actual capabilities of AI, and more to do with, as you've also likely noticed, a deep anxiety amongst investors fearing being 'left behind', or genuine belief in the odd promises of the increasingly bizarre proclamations by AI CEOs.
In the last entry, I discussed how in the aftermath of the 2007/08 financial collapse nothing really changed. There were no sweeping reforms of the system of the likes you saw in Britain and America in the aftermath of the 1929 collapse. There was a Milquetoast neo-liberal response, the kind of incrementalism and corporate welfare that we've all come to expect. 'The problem is too much regulation' decries The Economist, bringing out the World Bank's 'Ease of Doing Business Index' and saying that too many governments are making capitalist extraction too difficult. 'It's bureaucratic systems strangling innovation' swears The Financial Times, as it presents stories of brave Silicon Valley Venture Capitalists building 'the future' on the edge of government oversight, decrying how backwards the cabbies of London are for demanding that Uber be regulated in kind. The launch of the iPhone in 2007 had lined up with the economic collapse of 2007/08 to bring the perfect storm: Silicon Valley, not these inefficient governments, would save us from the ongoing financial collapse. As David Cameron and Barack Obama took their respective offices, they were more than happy to hand the keys for economic revitalisation to anything and everything tech-adjacent.
And thus we exchanged one bubble for another.
Ground breaking, world changing technology doesn't actually come that often, and when it does it's typically the culmination of lots of previous development. The Wright brothers were far from the only people working on the question of powered heavier-than-air flight, but their capacity to combine to the pre-existing and ongoing development of the internal combustion engine to the pre-existing and ongoing development of gliders was the breakthrough that many on 'both sides of the Atlantic' had been waiting for. The iPhone was not the first smart phone, nor the first device to use capacitive touch rather than pressure-based touch, but it was the first to combine them in a way ready for the mass market. Not to mention, the branding mixed it with previously successful Apple products, the iMac and the iPod. Lots of people forget how big the iPod was. Nonetheless, Steve Jobs (who had little actual work on the product designed by the large team of engineers he routinely abused) became a messianic figure in Silicon Valley. The once deposed CEO returned to his former company, changed the world, and made Apple the single most valuable company in the history of the New York Stock Exchange. Steve Jobs went from being a, by all accounts, deeply unpleasant man who didn't believe in personal hygiene, treating his daughter well, or generally doing most things we associate with being a good human being, to the man with all the answers for the future of humanity. He became the model CEO, and no shortage of Silicon Valley CEOs wanted to be the next Steve Jobs.
The problem, like I said, is that things like the iPhone are uncommon. Further, computers, like their creators and users, have to obey the laws of physics. We're reaching the physical limits of Moore's law, lithium-ion batteries have remarkably obvious downsides, but no one's worked out a better way yet. The internet is reaching a point of omnipresence that renders diminishing returns on profitability from further expansion. All 'innovation' is being lead by a handful of executives at a diminishing number of firms as capital and intellectual property are increasingly monopolised in the firms that are desperately trying to push the next big innovation out, now with the added pressure of being the thing that everyone is looking towards to save us, after a decade of propaganda painting big tech as just that.
Of course this was going to create a cult. And if you don't count capitalism itself as a cult, we're probably inching towards our first cult-based economic recession.
The drive to be the next Steve Jobs has already given us Sam Bankman-Fried, Elizabeth Holmes, the 'effective altruists' and now the 'effective accelerationists'. These are all close partners of each other. The 2007/08 financial crisis was never solved, the causes never went away, instead the increasingly dead weight of neo-liberal capital was placed on the shoulders of Silicon Valley with hope that it'd keep giving us new iPhones to stave off economic collapse, with Silicon Valley and the various governments all believing that Silicon Valley would be the future. The promise of AI becoming a new God to solve all these problems that we gave Silicon Valley to solve would, of course, give us the cult of the TESCREALists.
Dyer-Witheford (2025) talks about how, you guessed it, Marx saw something like this coming. Capitalism was always going to jump from crisis to crisis as the capitalist class sought to extract surplus value from 'fictitious capital'. The bourgeois class, Dyer-Witheford, Marx, and now me, argue, prefer fictitious capital to actual capital as it is remarkably faster to withdraw profits from.
Let's return to Apple for a second, since they became the source of our ongoing collapse. The foundations for Apple come from the Apple II line of computers. Rather famously in computing circles the Apple II was based on improvements to Steve Wozniak's Apple I. The Apple I had been available to the consumer, but as a kit, not unlike flat pack furniture, with the onus being on the buyer to build the computer at home. The Apple II's biggest obvious advancement over the I was that it was prebuilt, and needed only be plugged into a television and the mains to work. The Apple II wasn't the only 'microcomputer' readily available to consumers, it joined the ranks of the Commodore 64, the Sinclair ZX, the BBC Micro, the Acorn Electron, the original IBM PC (from whence we got the term 'PC'), and several, several, several other microcomputers. But the Apple II moved enough units to establish Apple as a household name. 'The Woz', himself a keen engineer had designed the Apple II with a sense of engineer. Apple's marketing department were keen to follow up the success of the Apple II with the Apple III.
If you're a retro computing enthusiast, like me, you've come across some variation of the Apple II. I have an Apple IIe in my personal collection. Apple II's are extremely common because they were extremely well built and built a considerable user base. Apple III's are measurably harder to come across. Largely because they were the first casualty of Steve Jobs' war on reality. Steve Jobs, like most capitalist, hated objective reality whenever he felt it got in his way to extract surplus value from labour. For reasons known only to Jobs, he had convinced himself that the noise from computer fans were deterring consumers from buying computers. As such, it was vital to build a 'silent computer' that was devoid of fans. I don't mean to brag, but I was pretty good at physics in secondary school and sixth-form, so let me bring you up to speed on the laws of physics:
All energy is expressible as heat
Electricity is the form of energy transfer that powers a computer
Electricity is a form of energy
Electricity, therefore, is often expressible as heat.
Computers, you may have noticed, run hot. Leading to us needing to cool them. The Apple III rather notoriously would heat itself to death. It was not the last Steve Jobs 'designed' product to run into this problem. The Lisa, the original Macintosh, and the infamous G4 cube all routinely ran into cooling problems. Granted, none of them hit the thermal wall nearly as hard as the Apple III did, but they all showcased Jobs' desire to assert his will over the literal laws of physics, with unsurprising results. Jobs' war with objective reality would, rather poetically, kill him outright when he attempted to cure cancer with fruit juice. Nonetheless, his canonical position of 'Silicon Valley C-Suite Jesus' insured that the war on reality would continue, despite the most obvious casualty.
This is, of course, because capital itself frequently defaults to this war with objective reality. It has to, that's the only that promises of infinite, constant, and quarter-on-quarter growth is even possible. Here we get to the biggest thing that detracts the capitalist mode of production from previous expressions of the economy. There has to always be another iPhone in the wings, even if that's physically impossible, and frankly it is.
The 'real' productive cycle is relatively simple, capital seeks to make a physical good. It spends a variable amount of capital to purchase the commodities (including commoditised labour) to produce these goods, then sells them to receive the surplus value. The actual surplus value is derived from the 'invisible gift' that the human labour provides in this process, even if that gift isn't immediately obvious to the labour or the capitalist. In fact, the capitalist seeks to provide a layer of abstraction to the labour in order to mask the value of labour from labour, so they don't do anything pesky like organise. Indeed, alienating labour from their value is part of the wider process capital undertakes in order to minimise costs and maximise profits. Capital would love to figure out a way to derive new capital from existing capital without needing to purchase any commodity. As Dyer-Witheford puts its, capital would love to go from M to M' without having to invest in commodities (C). This constant desire of capital to derive surplus value without investment in any commodities is historically present in the stock market or in deriving interest from loans, but even this still called for a certain amount of investment in commodities. Nonetheless, the immense amount of capital generated through simple speculation, the form that allowed the least amount of investment, had resulted in a system already driven largely by 'fictitious capital' as early as the 19th century.
The levels of abstraction that computers and the internet afford the capitalist class has allowed for even more capital circulation to generate 'fictitious capital' than their 19th century counterparts could ever dream of. The spectre of AI allows for a level of abstraction from the value of labour that would make Henry Frick drool, as Silicon Valley has used the promise of a 'God computer' to alienate the value of everyone from medical doctors to local governments.
Silicon Valley's biggest problem, of course, is that like many spectres, AI is an elaborate party trick. The thinking, reasoning, 'God computer' has as much practical reality behind it as does 'the Wizard' in The Wizard of Oz. But still they push the narrative, complete with trying to convince us of a choice between a benevolent 'New Testament' God versus the wrathful 'Old Testament' one, to forever drive up share holder value. Extracting as much real capital as possible before the 'market corrects' and the fictitious capital does what it does best, and disappears. None of the prevalent AI pushers will suffer any consequences, but billions of people who had no real stake in the matter will suffer immensely as we once again discover that a handful of corporations were dangerously overvalued. Silicon Valley will only survive if they can find another bubble to push. But frankly, they're tapped out. The reason they're pushing AI as hard as they are is because there's no real 'next iPhone' in the wings. They don't have another big trick to cover up the ongoing stagnation and failure of capitalism. Consumers can no longer afford to consume, and the system is continuing the rot its been experiencing since 2007/08. The Silicon Valley distraction post-07/08 just being a macabre Weekend at Bernie's with the bloated corpse of neo-liberal economics.
Silicon Valley's prominence as the holders of the future will fade like the car manufacturers before them, and for the same reason, you can't consumer products your way out of capitalism's inherent flaws. Capital needs labour, but labour doesn't need capital. We can, should, and probably will move on from capitalism. Let's just hope it's too a better system. Labour predates capital, and will outlast capital. Maybe it's time we let labour own the means of production.
'Workers of the world unite', and all that.
Join the conversation