An AI Reckoning, When the Mag-7 Forget History

So the momentum picks up and pundits are betting on the AI bust … ‘The Big Short’s’ Michael Burry is back with cryptic messages and two massive bets, you guessed it, Mag-7 members are up front and central as prospects in a AI bubble collapse. So the following thoughts maybe a bit academic if this AI Boom does transpire to be a dot.com re-run. As I speculated in ‘Is The AI Boom just the Birth of New Infrastructure?‘, we could in fact be witnessing the creation of the world’s next great layer of infrastructure, one that, once commoditised, becomes the fertile substrate for the next century of innovation, or not!

One think is for sure, central banks and governments will do what they can to keep the AI bubble boyant, they cannot afford to lose the momentum it is giving stocks and the economy. Reflet on the fact that AI investment has and remains the primary source of America’s economic and stock market outperformance. Without it, stocks and GDP would be looking very different indeed and as Ludwig von Mises the economist famously said you cannot inflate a bubble that has popped. Meaning that central banks cannot rescue the AI bubble if it really has a mortal wound after this week’s correction, but based on past behaviour it will not stop them from trying at all our expense.

I digress, so pivoting into my latest exploratory thoughts on the theme …

IF this AI boom does maintain its cadence, even the central bankers and governments cannot avoid observing it is starting to feel uncomfortably familiar. The same companies I wrote about in A Hidden Faultline in the AI Boom, that once claimed to ‘democratise’ technology now dominate the AI era like a digital cartel. Microsoft, Nvidia, Apple, Alphabet, Amazon, Meta and Tesla, the Mag-7, have woven themselves into an incestuous lattice of cross-investment, dependency and co-marketing that increasingly resembles a self-referential cabal rather than a competitive market. Each funds, fuels, features and even champions the others’ AI ambitions, risking locking smaller innovators out and drawing passive capital into a narrow, over-inflated segment of the S&P 500.

History offers plenty of warnings. Remember the Baby Bells? Then there was IBM, Standard Oil, AT&T the dot-com monopolies of the 2000s, every era of technological revolution has flirted with dominance before guess who arrives … yup, the regulators. The Mag-7 appear to have learned little from those lessons. Their alliances may accelerate innovation today, but they also concentrate systemic risk, economic, political and even algorithmic, in the hands of a few. Sooner or later, the world’s monopolies commissions will catch up and the fall from grace could be brutal with damaging collateral impact on an unsuspecting ecosystem of SMEs (Small Medium Enterprises) and businesses that are inevitably becoming deeply coupled to AI services. AI is becoming like the new digital crack cocaine, from chatbots to recommendation engines, they deliver instant intellectual gratification to  individuals and businesses alike. Its accelerating adoption mirrors the arc of any powerful stimulant, seductive, performance-enhancing in the short term, but corrosive to independence and resilience if unmoderated. Hmm… let’s leave that thought exercise for another missive …

Back to the regulators. Would it be too much to hope for a modern approach that could break this pattern? Rather than doubling down on vertical control and exclusivity, the AI industry should pivot toward federated innovation, open architectures, transparent AI governance, interoperable standards and shared model assurance frameworks. Build a collective new superstructure of TRUST, leaving behind the invidious social media industry trend of hooking users like some esoteric quarry. True competitive advantage in this new era will not come from hoarding GPUs (Graphics Processing Units) or data, but from trust, transparency and ethical velocity, the ability to scale AI safely, collaboratively and with societal licence. If the Mag-7 do not modernise their mindset, they will discover that regulators and markets have a long memory. This could also facilitate a softer landing in the event of an AI bust scenario and be in the economic best interests of all concerned.

For now, the Mag-7 march on, hand in silicon hand, chanting the gospel of open AI while quietly patenting every bolt that holds the temple together and ties in subscribers. They promise to empower humanity, yet somehow humanity keeps ending up as the unpaid intern, the digitally enslaved to corporate agendas. When the antitrust dawn finally breaks, we will watch these titans solemnly testify that no one could possibly have foreseen such concentration, right before they each spin up a new subsidiary with names that inflect the likes of ‘FairCompete.ai’. History does not just repeat itself in tech; it reboots with shinier graphics and the same patch notes and of course vulnerabilities and privacy compromises abound. Chancers, some never learn …