AI reckoning as a term is a rather stark warning. Even more so when it comes from Aswath Damodaran who is a professor of finance at the Stern School of Business at NYU (New York University).
The reckoning that Damodaran is concerned about is a ‘Minsky moment‘ and represents a stark contrast to the boundless techno-optimism of Marc Andreessen.
The AI reckoning wasn’t on the mind of investors who bought into Cerebras Systems.
Cerebras Systems makes wafer sized chips with memory and processing on the same die. This reduces latency and increases speed. As cool as Cerebras Systems technology is, the company currently has sales of $510 million and was valued on its opening day of trading at $71 billion.
AI-Powered Cyberattacks | Robert Scoble and A glimpse into cyber-security’s AI-driven future | The Economist – A few years ago a participant used the conference network to hack a water-treatment facility in America (Messrs Wyler and Stump are cagey about the details). Another hid behind the din of legitimate hacker traffic to attack government websites and payment systems. The noc team traced him, sent him a message reminding him that doing illegal things from Black Hat was still illegal, then watched him close his laptop and walk away. Hackers on the other side of the world try their luck too. When the registration server was switched on, attacks began at once, including traffic that appeared to originate in Romania….Mr Stump says the noc has seen a pattern across multiple Black Hat conferences in which Taiwanese participants show up with hacked devices. “Most of [the traffic] goes back to China,” he says. ai-powered attacks by nation-states or cybercriminals are likely to intensify.. The team thinks the ai race is only beginning. For Mr Wyler, the vulnerabilities discovered by Mythos, including some that have gone undetected for decades, are to be welcomed rather than feared. “We now know they’re there.” All the same, cautions Mr Stump, the next two years will be turbulent, as more flaws will be uncovered; more breaches will occur as firms feed sensitive data into ai systems; and more insecure code will be written.
My love of films by Wong Kar wai came at a time of changing media. In the early 2000s I watched the films the first time after I got a portable multi-region DVD player and there had been a massive surge in video labels including Artificial Eye and Tartan publishing arthouse titles. This provided a great cost-effective way to explore and experience world cinema and I gravitated towards Japanese and Hong Kong cinema.
I was already familiar with traditional martial arts films and the ‘gun fu’ of John Woo. Wong Kar wai was Hong Kong’s answer to French new wave auteurs.
Around the same time, I ended up dating someone who lived in Hong Kong when we bonded over Faye Wong’s performance in Chungking Express. In a moment of delicious irony, I got to watch Wong Ka wai’s one ‘western’ film My Blueberry Nights while staying in Hong Kong.
While Wong Ka wai’s filmography wasn’t the reason why I moved to Hong Kong, but it was a reason why I moved to city and had the privilege of living there for a while.
This time around I was working my way through Criterion’s World of Wong Kar wai boxset which had been bought for my birthday during COVID time.
The Wong Kar wai boxset contains:
As Tears Go By
Days of Being Wild
Chungking Express
Fallen Angels
Happy Together
In The Mood for Love
2046
As Tears Go By
As Tears Go By was released in 1988. It is one of Wong Ka wai’s most conventional films from a Hong Kong perspective. Andy Lau plays the protagonist Wah, a triad soldier. Wong borrowed from the plot line in of Martin Scorsese’s 1973 film Mean Streets in terms of the story revolving around dynamic of two friends, one of whom is irresponsible. It’s a great stylish film, but if you told me that it had been made by Ringo Lam, Tsui Hark or Johnnie To, I’d have believed you. Hong Kong cinema audiences loved it and it would be another 25 years before Wong made another film as popular with local cinema goers.
Days of Being Wild
1990’s Days of Being Wild is often considered by some to be part of an informal trilogy, the others being In The Mood for Love and 2046. Stylistically it features love, loss, similar pacing, the use of inner narratives and experiments with colour. Thematically they are sensitive to the passage of time and have a vice-like hold on emotional memory. But the threads between the characters aren’t really bought together until the next film in the trilogy. Tony Leung’s character and is given the name of ‘Gambler’ is only wordlessly introduced right at the end of Days of Being Wild. He only becomes known as Chow Mo wan in the next film: In The Mood for Love.
Leslie Cheung was ideally cast as a lost, rootless, self-absorbed drifter Yuddy in the film. Yuddy has a hole at the centre of his being that he is unable to fill. Like many dislocated Chinese in Hong Kong between the civil war and the cultural revolution he drifts.
Yuddy is described as a legless bird, only touching down with death. Critics have interpreted this as pre-1997 handover anxiety. An article published on the Hong Kong Film Critics Society website described it for me best
At the time, they were echoes of Hong Kongers’ sentiments under the looming 1997 deadline. Leslie Cheung, who has a love-hate relationship with his foster mother and after a failed mission to find his biological mother, drifts in self-imposed exile, is a metaphor for the city caught between the two sovereign states of China and England.
Set in the 60s, the film is filled with signifiers of nostalgia (props, costumes, music, scenery). Reminiscence is but a lament that the good of the present will not last. And Days of Being Wild is but an elegy for a Hong Kong caught between 1989 and 1997.
The film follows Yuddy and the trail of emotional wreckage he leaves in his wake as he looks to track down his biological mother.
Secondary threads follow Su Li zhen, played by Maggie Cheung and Carrie Lau’s performance as Leung Fung ying. Both of whom where Yuddy’s transient love interests.
There is a scene where Yuddy looks to obtain an American passport and something about Cheung’s movement and the casual violence reminded me of Michael Madsen’s Mr Blonde in Reservoir Dogs released three years later. While Mr Blonde is nihilistic, he lacks the duality of Cheung’s character.
Leslie Cheung gives a fantastic performance on screen and his tragic death 13 years later was a serious blow to Hong Kong film-making.
Days of Being Wild wasn’t well received by Hong Kong cinema goers at the time, despite being well regarded by film critics everywhere. This adds to Wong Kar wai’s reputation as an auteur streets ahead of the local audience. Which was a view I bought into the first time around.
Having watched it again years later, I have some hypotheses as to why it didn’t do well.
While the dislocation and drifting reflected life for many in early 1960s Hong Kong, it didn’t match the go-go economy and Lion Rock can-do spirit of Hong Kongers in the following three decades. Younger audiences wouldn’t be able to relate to it in the same way. Cinema audiences tend to be younger than the general population, so that disconnect makes a degree of sense.
After the Sino-British joint declaration was signed in 1984, a pre-handover anxiety hung over Hong Kong. KMT supporting newspapers gradually closed down or pivoted their editorial style. Astronaut families became commonplace with upper middle class children based outside the city in Vancouver or Australia while their parents made money in the run up to handover due to the go-go environment of the time. The second passport, gave the family a bolt hole in case things went wrong after China took over. Local Hong Kong cinema goers wanted the escapism of action films, gambling movies and comedy.
It is very different in pace as a film compared to its high-octane peers at the time from the likes of John Woo. This time I got to see the original Hong Kong trailer of Days of Being Wild – and could understand how you could go into the cinema expecting something with much more pace rather than the dream-like experience much of Days of Being Wild gives you.
The ending came abruptly and without context.
Chungkung Express
1994’s Chungking Express is two locations and views of modern Hong Kong. Trading hub Chungking Mansions in Tsim Sha Tsui and the Midnight Express takeaway restaurant in Lan Kwai Fong. While it is a romantic comedy of sorts and varies in pace, it also has the qualities of what we now expect in a Wong Kar wai film. The focus on time, distance and emotional memory – but with a much lighter touch than Days of Being Wild.
The film is anthology of two stories with retail worker Faye being the one serendipitous point of connection that holds both stories together.
Takeshi Kaneshiro plays alongside fellow Taiwanese actress Bridget Lin in the first story. He is estranged from his girlfriend May relying on calls to her parents and tins of pineapple which is her favourite food.
In the second story tells of how Cop 663, played by Tony Leung is left by his air stewardess girlfriend and acquires a stalker while getting over his old love.
Both stories caught the urban energy of 1990s Hong Kong and resonated better with audiences. What is more remarkable is how fast the film was made in an improvisational way with guerrilla film making techniques. Its looseness was by design as Wong tried to mirror Haruki Murakami’s writing style on screen. The film was shot in a two-week break from editing Ashes of Time.
The film brought Faye Wong to an international audience and cemented Wong Kar wai’s arthouse credentials.
The film feels very now in terms of its style and even the use of old technology like camcorders and pagers on screen doesn’t pull you out of the film in the way I might have expected.
Fallen Angels
1995’s Fallen Angels was a surprise to me the first time I watched it. In some ways it goes back to As Tears Go By in its exploration of Hong Kong’s organised crime world. You have Wong Kar wai’s use of loose narrative, colour, tight spaces and urban energy.
Compared to As Tears Go By it’s a slower paced film. It makes what now comes across as innovative use of close up wide angle photography that makes it look as if its shot on an iPhone decades before the modern iPhone came out. It also feels computer game-like in the action sequences. I can also understand why it has been compared to music videos in terms of style. Parts of the action sequence reminded me of Point Break in terms of the camera viewpoint. While it feels ‘intimate’ because of its claustrophobic shots, local cinema audiences didn’t relate to Fallen Angels.
There is a duality to the film. Takeshi Kaneshiro plays Police Officer Ho Chi moo badge number 223 in Chungking Express. In Fallen Angels he is Hoo Chi moo, convict number 223.
There are wider Wong Kar wai touches, in mid 20th century artefacts from 1960s Hong Kong architecture to the Enicar illuminated wall clock in the assassins base.
(Enicar was a historic Swiss watch brand that was best known across Asia and China. The brand name is now owned by Wah Ming Hong who had been their distributor in China since the 1930s. One of my first memories of Hong Kong was giant building wrapping adverts for Enicar and another former Swiss, now Hong Kong watch company Solvil et Titus).
Watching it this time, I had a nagging feeling that something had changed and sure enough when I searched online. I found out that the film had been extensively cropped shot by shot and recoloured by Wong Kar wai in 2020, not always for the better.
Happy Together
Happy Together was released in 1997 and is still a highly regarded example of New Queer Cinema alongside the like of Querelle.
As a work out of Hong Kong it’s remarkable. Hong Kong as a society is conservative and there is a don’t ask, don’t tell aspect to the treatment of the LGBTQI community in Hong Kong. Legislation is more advanced than Hong Kong society at large.
Leslie Cheung was an ideal protagonist known for being a champion of the avant-garde and having on-screen characters that experimented in different forms and levels of masculinity.
Happy Together‘s themes of displacement, exile, and the repeated line “Let’s start over” mirrored loneliness, heartbreak, the collective uncertainty and “fretful wanderlust” of Hong Kongers at the time. The film came out in Hong Kong just two months prior to the handover of Hong Kong to China.
Happy Together was released with a Category III rating, meaning only those 18 and above can watch it. Primarily this was down to Hong Kong’s greater latitude for violence rather than sex on-screen. Wong Kar wai’s eye and treatment on camera means that Argentina feels rather like Hong Kong in the film.
In The Mood For Love
For many people, 2000’s InThe Mood For Love is the gateway drug to Wong Kar wai films. I think that the Cantonese title ‘Flower-like Years’, ‘the prime of one’s youth’ suits the film better. It’s the second film Wong’s informal trilogy, after Days of Being Wild.
The contrast in pacing between In The Mood For Love and the previous film mirrored the move from frantic kinetic energy of colonial Hong Kong living on borrowed time to an oppressive study in stillness focused on longing for the past post-handover. Wong admitted in interviews that the hotel room number, 2046, where the protagonists get together to write represented the final year of Hong Kong’s ‘guaranteed‘ autonomy. The Shanghainese dialect spoken, outfits and food reflected Hong Kongers nostalgia for the post-civil war era of migration to Hong Kong as time of pain and hope.
We are properly introduced to Tony Leung’s Chow Mo-wan character. Su Li zhen, played by Maggie Cheung adds the real line of continuity. We know it’s 1962, and probably at least a few years since Days of Being Wild. On one level their fortunes have improved, Su now works as a secretary for a shipping company; Chow is a journalist. They are both unhappily married and find solace in each other’s company.
It’s like David Lean’s Brief Encounter but with the colour and latent emotion dialled up through copious amounts of hallucinogen. The loneliness, missed connections, the weight of time, regret, longing and rootlessness feel even more intense in this film than any of Wong’s previous films. The also a sharp contrast with the licentious and violent elements in Wong’s previous films.
I remember watching it the first time and being blown away by it visually without taking in the plot, performance and nuance layer throughout. I then revisited my old DVD copy several times later on.
Wong lays out their collective journey of discovery in finding out that their respective partners are having an affair. This builds the closeness and tension them, as does the martial arts serial that they write together.
(This always struck me as a nod the popularity of Hong Kong based authors like Liang Yusheng and Louis Cha who worked as newspaper journalists, before going on to write serials published in newspapers and magazines. Eventually their works would be adopted in Hong Kong films including Wong Kar wai’s own The Eagle Shooting Heroes and Ashes of Time; and TV series in Hong Kong, Taiwan and communist China).
In The Mood For Love redefined the way male main actor roles were portrayed in Hong Kong cinema allowing greater character depth than was previously the case with gun fu, wuxia and action comedies. It gave the post-handover Hong Kong film industry a much-needed creative shot in the arm before the ‘China-Hong Kong’ joint ventures finally bled it dry.
In The Mood For Love seemed to be the ground zero for Hong Kong mid 1960s nostalgia, such as the G.O.D ‘Bing Sutt Corner’ redesign of the Starbucks branch on Duddell Street in the central district of Hong Kong. Others got in on the act, 7-Eleven released a set of ‘Old Hong Kong’ phone charms.
Writing this post, I went back to find out what I had written about the phone charms.
There is a wider trend of nostalgia in the city which 7 Eleven Hong Kong is tapping into.
It is interesting because it reflects a widely held view that the bright new future offered by mainland China isn’t bright, attractive or desirable. This will likely cause trouble in Hong Kong for China in the future; if it rolls out from the cultural zeitgeist into political aspects of Hong Kong life.
There are times when I wish I was wrong.
Nothing jarred from memory when I rewatched In The Mood For Love, but I had forgotten the documentary footage of President De Gaulle visiting Cambodia near the end.
2046
After In The Mood For Love, 2046 follows Chow Mo wan as he attempts to get Su Li zhen out of his system. The story also connects with Days of Being Wild with Carina Lau’s character still being heart-broken over the death of Yuddy, years later.
There is a line in Tony Leung’s monologue that encapsulates 2046 the central plot premise really well.
“Love is all a matter of timing, it’s no good meeting the right person too soon or too late. If I’d lived in another time or place then my story might have had a very different ending” – Tony Leung’s character Chow Mo-wan in 2046
2046 captured post-Handover disillusionment, a community that realises its own ephemeral nature. Hong Kong’s specialness appears as suffering according to Stephen Teo – and I think he got it right.
…a visually ravishing work that’s downright apocalyptic in its suffocating sense of dread and despair. – David Pountain, Little White Lies
I finished the boxset, drained in a good way, but also disappointed, not in Wong Kar wai’s work but in the Hong Kong it now exists in. As it was once my home, I felt broken.
Marc Andreessen’s 2026 AI outlook was published by A16z. As one of the leading funder of Silicon Valley startups, his world view matters.
I’ve gone through and contrasted his 2026 AI outlook through the lens of the viewpoint I researched and wrote up. The core point over where we differ: Andreessen see’s the dawn of a new unbound industrial revolution. Whereas I think that the upside he sees is hard to navigate to; and also risk mirroring the echoes of financial bubbles past in many of the high profile pure play AI companies like OpenAI or C3.ai.
Andreessen has been the quintessential techno-optimist for at least the past two decades with his emblematic essay Why Software Is Eating the World. His worldview is of an industry where demand is insatiable and revenue is undeniably real.
Like Andreessen in his 2026 AI outlook, I would agree that players like Alphabet, Amazon and Microsoft are making money to variable degrees from their AI offerings as part of cloud computing and productivity software bundles.
But in my view there are two aspects of concern:
Pure plays with the notable exception of Anthropic don’t seem to have a clear path to profitability in a time period that would match their implied future revenues based on loans and valuations.
Hyperscalers like Meta and Microsoft are using unusual partnerships to fund their infrastructure and have been inconsistent over how fast they would depreciate their AI computing hardware.
Both Mr Andreessen and myself agree that we are witnessing a technological shift of seismic proportions, arguably, “bigger than the internet” as he put it.
I think that change will happen slower in the short term due to an economic sand box acting as a rate limiter on infrastructure and derived labour efficiencies. In particular, the economic viability of the infrastructure being built to support it.
The Nature of the Boom: Real Revenue or Irrational Exuberance?
A key question is the the solidity of the current market.
For Andreessen, the AI boom is not a speculation fuelled exercise but a demand-driven reality. He argues that unlike the early internet, which required years of physical infrastructure build-out before finding business models, AI is generating cash immediately.
“This new wave of AI companies is growing revenue like… actual customer revenue, actual demand translated through to dollars showing up in bank accounts at like an absolutely unprecedented takeoff rate.”
Andreessen points to “revealed preferences”, observed insight from what people do is more insightful than what they say when asked by market researchers. While the US and European publics express fear of job losses, they are simultaneously adopting AI tools at a fast pace.
My “Dot LLM Era” report, however, suggests this revenue may be dwarfed by the capital expenditure required to generate it.
It posits that the sector faces a “self-defeating economic” cycle. The report highlights that the current valuations of the “Magnificent 10” tech giants (including Nvidia, Microsoft, and Alphabet) imply a forward price-to-earnings (P/E) ratio of 35 times. This is alarmingly close to the S&P 500’s P/E ratio at the peak of the dot-com boom, which approached 33.
In the report I warn that current valuations assume LLMs will drive revenue growth by $1–4 trillion in the next two years.
If this growth is achieved through massive job automation, the resulting unemployment could depress the very economy needed to sustain these companies. Like the 2008 financial crisis, this would invite various forms of regulatory intervention.
This implies a sweet spot between speed of productivity gains versus efficiencies in terms of job losses realised by clients – and acts as a break on the velocity of adoption.
The Infrastructure Trap: The Amortisation Crisis
Perhaps the most critical technical divergence between our viewpoints concerns the hardware powering this revolution.
The “Dot LLM Era” report introduces a chilling concept: Amortisation Risk. It draws a comparison to the “telecoms bubble” of the late 90s, where companies laid massive amounts of fibre optic cable.
The Telecoms Analogy: Fibre optic cable had a useful life of over a decade, meaning even after the companies went bust (like WorldCom), the infrastructure remained useful for Web 2.0.
The AI Reality: Modern AI infrastructure relies on GPUs and TPUs. These processors have a useful life of only 3 to 5 years before becoming technically obsolete.
The report argues that if an AI bust occurs, the hardware will not be waiting for a resurgence; it will be electronic waste.
Hyperscalers are currently lengthening the assumed useful lives of this hardware in their financial filings—Google to 6 years, Meta to 5 years—which the report suggests may artificially overstate profits and adversely affect competitors with debt securitised against AI data centre hardware.
The rapid obsolescence of chips represents a “financial and technological amortisation risk” that could lead to trillions in write-offs, similar to the $180 billion loss left by WorldCom.
Andreessen views this hardware cycle through a different lens: Elasticity.
He acknowledges that massive investment leads to gluts, but argues that “the number one cause of a glut is a shortage”. He believes the price of AI compute is falling “much faster than Moore’s Law”.
As prices collapse, demand will expand exponentially. If chips become cheap and plentiful, AI can be embedded into everything, moving from massive “God models” in data centres down to small models running on local devices.
Geopolitics: The US-China Race
Both Andreessen and I agree that the AI landscape is a bipolar contest between the United States and China, but their assessments of the leaderboard differ.
Andreessen frames this as a “new Cold War” where the US must maintain dominance. He is encouraged by the “DeepSeek moment”, referring to a powerful open-source model released by a Chinese hedge fund. To him, this proves that catching up is possible and that the US cannot rest on its laurels. He advocates for open source as a way to proliferate American standards globally.
I offer a more sobering assessment of American pre-eminence. I argue that unlike the 1990s “Long Boom,” where the US was the undisputed hegemon; the current era is defined by high debt and strong competition. The US maybe on the Soviet side of a Reagan era ‘Space Defence Initative (aka Star Wars)’ AI race from an economic perspective.
Alibaba’s Qwen model claims to deliver comparable performance to American models while requiring 82% fewer Nvidia processors to run.
China doesn’t have a electrical power crunch in the same way that western data centres have due to its sustained investment in coal, nuclear, gas and renewable power sources.
Even Silicon Valley investors and major companies like Airbnb are opting for Chinese open-source models because they are “way more performant and much cheaper”.
Andreessen worries about regulation stifling US innovation, I think that the real geopolitical threats are:
China’s ability to operate largely immune to US sanctions, smuggling chips and utilising data centres in neutral geographies like Malaysia.
China’s speed at propagating the use of AI within its own populace, driving utility at a lower cost per token than their US competitors with state regulation defining trial ‘sand pits’. This will drive new uses faster in China and in China’s client states across the global south.
The Outcome: Transformation or “Minsky Moment”?
Where is this all heading? Andreessen is betting on a future where AI becomes as ubiquitous and essential as electricity. He envisions a “pyramid” structure: a few massive “God models” at the top, cascading down to billions of specialised, cheap models running on edge devices. He admits these are “trillion-dollar questions,” but his firm is aggressively investing in every viable strategy.
I think that LLMs will make a sustained technological impact, but it will be limited in velocity by economic boundaries. In the “Dot LLM Era” report, I outlined seven potential scenarios, ranging from total transformation to total collapse. Currently, it assigns a ~95% likelihood to the “Moral Hazard” scenario.
The Moral Hazard: This scenario posits that major AI players will be considered “too big to fail” due to national security imperatives. Governments will step in with loan guarantees and subsidies to backstop the massive infrastructure debts, effectively nationalising the risk . Rather like the banks during the 2008 financial crisis.
The Telecoms Bust: With a ~75% likelihood, the report fears a “Minsky Moment”, a sudden market collapse driven by the realisation that cash flows cannot cover the massive debts incurred to build short-lived data centres.
Comparative Summary of Perspectives
Feature
Marc Andreessen (The techno-optimist)
My own view (The economic limiting skeptic)
Current Phase
Inning 1. “Biggest technological revolution of my life.”
Phase 1: The boom / The Inflation of a bubble.
Revenue
Real, unprecedented, showing up in bank accounts.
Potentially illusory for at least some players; reliant on untenable cost savings.
Infrastructure
Shortages lead to gluts; cheap chips drive adoption.
Amortisation risk: hardware obsolescence in 3-5 years.
Pricing
Usage-based is great for startups; prices falling fast.
US dominance challenged; Chinese models are more efficient and effective enough for organisations from Singapore to Silicon Valley to adopt them.
Outcome
Long-term ubiquity; widespread prosperity.
Technological change bounded by economic limitations. Current high risk of “Minsky Moment” or government bailouts.
Conclusion: The Trillion-Dollar Questions
Marc Andreessen candidly admits that “companies… need to answer these questions and if they get the answers wrong, they’re really in trouble”. His firm’s strategy is to bet on everything: large models, small models, apps, and infrastructure. He is doing this on the assumption that the aggregate wave will lift all boats.
My ‘Dot LLM Era’ report offers a counterweight to this enthusiasm. A bubble decouples technological and financial progress. Technological utility does not always equal investor profit. As I note in ‘Dot LLM Era’, “Bubbles don’t kill technology from moving forwards”. The internet did change everything, but it also wiped out trillions in shareholder value along the way.
The defining question for the next five years is whether the demand for AI can grow fast enough to pay for the hardware before that hardware becomes obsolete.
If Andreessen is right, this elasticity of demand would save the day. If I am right, we may be heading for the most expensive recycling project in human history.
The Productivity Paradox and Society
A fascinating tension exists between Andreessen’s view of societal adoption and my own macroeconomic warnings regarding productivity and labour.
The “Wingman” Economy vs. The Phillips Curve
Andreessen describes a “symbiotic relationship” where AI acts as a productivity multiplier: a “wingman” for doctors, coders, and writers. He argues that higher pricing in SaaS can actually benefit the customer by funding better R&D, suggesting a cycle of value creation.
I argue that the wingman sweetspot is optimal but tricky to land, in the report I showed the risk through a darker macroeconomic “thought experiment.” It used the Phillips Curve and Okun’s Law to model what happens if Andreessen AI scenario succeeds too well.
The Thought Experiment: If AI automation generates $1 trillion in cost savings through job cuts, it implies approximately 10.5 million unemployed US workers.
The Consequence: Such a spike in unemployment could trigger deflation and a massive drop in GDP. This is “self-defeating economics”: the hyperscalers need a healthy economy to consume their services, yet their success might undermine the investor, enterprise customer and consumer base.
Andreessen counters this fear by citing historical context. He notes the “Committee for the Triple Revolution” in 1964, warned Lyndon B. Johnson that automation would ruin the economy—a prediction that proved false. He believes AI will follow the path of electricity or the internet: initially terrifying, eventually indispensable.
Automation did displace a massive amount of developed world jobs moving at a much slower pace than Andreessen predicted for AI. Electricity moved at an equally slow pace compared to the pace envisioned by AI’s champions.
The Open Source Debate
Both of us agree that the role of open source is pivotal in both narratives.
For Andreessen, open source is the great accelerator. He marvels at how knowledge is proliferating: “Some of the best AI people in the world are like 22, 23, 24”. He views the leak of knowledge as inevitable and beneficial for US competitiveness, provided the US stays ahead.
I analysed the “Red Hat Analogue.” It suggests that in the dot-com era, open-source (Linux) won, but the companies that built the models (or distributions) mostly failed, with Red Hat being the notable exception.
I assigned a ~70-80% likelihood to the “Red Hat Model,” where pure-play LLM creators (like OpenAI or Anthropic) might struggle to justify their capital burn as open-source models like Meta’s Llama or Alibaba’s Qwen commoditise the intelligence.
We have already seen Singapore’s national AI programme drop Llama for Alibaba’s Qwen, reinforcing the idea that the value might accrue to those who service the models, not those who create them.
Final Thoughts
The divergence between Marc Andreessen and my own analysis is not about whether AI works both of us would agree that the technology can be magical and transformative. The disagreement is about who pays for it, how that affects the velocity of AI andwho profits.
Andreessen sees a future of abundance where falling prices drive infinite demand.
My own view sees a future of financial reckoning shaping the 2026 AI outlook where shorter hardware lifespans and brutal competition erode margins, setting a slower pace at which we reach Andreessen’s abundance.
Andreessen’s viewpoint reminded me a lot of mid-20th century aspirations for nuclear power. Nuclear power offered a similar vision in the mid-20th century of electricity too cheap to meter. That was never close to being achieved in the likes of France – arguably the most passionate adopter.
As with the railway mania of the 1840s or the optical fibre boom of the 1990s, society may inherit a significant infrastructure, with a shorter lifespan built on the ashes of investor capital.
Our differing views boil down to a question for the 2026 AI outlook: are we in the “boom” phase, or are we staring down the barrel of the “amortisation crisis”?
As Andreessen himself concluded, “These are trillion-dollar questions, not answers”.
A post on AI sovereignty came out of one of those times when a casual conversation suddenly has you seeing the theme in your news feeds. I was having one of them conversations with a friend over a paper cup of coffee, mentioned I’d been embedded at Google and they said ‘we can’t trust the Americans with AI, the way we did with social’.
That opens opportunities. Chinese open source models are working in Singapore government data centres, Korean cloud computing company Naver is looking beyond its own country for clients who want an alternative to US big technology. France has gone it alone with its own defence AI – as the ultimate expression of AI sovereignty.
The All-Star Chinese AI Conversation of 2026 | ChinaTalk – Interesting discussions on China based AI platforms on their successes and challenges. By their nature, the give China defacto AI sovereignty. Risk taking and GPUs or TPUs performance seem to be the main sources of concern. A good deal of focus on squeezing out the maximum intelligence per watt rather than scaling to infinity and beyond. Tonality wise it’s refreshing down to earth in comparison to Altman et al.
‘South Korea’s Google’ pitches AI alternative to US and China | FT – Korea has built up positive relations in the Middle East since the 1970s when they helped on major construction and engineering projects. They would be viewed positively and as a good hedge to both the US and China from a technology dependency point-of-view. Their offer is greater AI sovereignty for Middle Eastern countries in particular, you might also winning business in Central Asia as well.
It comes as a growing number of brands are moving into the children’s, teenage and young adult skincare market. In October, the first skincare brand developed for under-14s, Ever-eden, launched in the US. Superdrug has just created a range for those aged between 13 and 28.
A number of brands have surged in popularity among very young social-media users, creating a phenomenon known as “Sephora kids”. These children share videos showcasing beauty products from Drunk Elephant, Bubble, Sol de Janeiro and similar brands.
A Theory of Dumb: Why Are IQ Scores Suddenly Falling? | Intelligencer – a century ago, if you asked someone what dogs and rabbits have in common, they might answer “Dogs hunt rabbits,” not “They’re both mammals.”Maybe, then, all the noise and novelty wasn’t rotting our minds but upgrading them. (Studies suggest that better nutrition and reduced exposure to lead may have also helped.) In any case, the Flynn effect held steady for so long and through so many apparent threats that there was no reason to believe it wouldn’t last forever, even if, someday, somebody invented a chatbot that could do homework or Theo Von started podcasting.
Or so thought Elizabeth Dworak, now an assistant professor at Northwestern University’s medical school, when she chose the topic of her 2023 master’s thesis. She decided to analyze the results of 394,378 IQ tests taken in the U.S. between 2006 and 2018 to see if they exhibited the same climb. “I had all this cognitive data and thought, Hey, there’s probably a Flynn effect in there,” she says. But when she ran the numbers, “I felt like I was in Don’t Look Up,” the movie in which an astronomy grad student played by Jennifer Lawrence discovers a comet speeding toward Earth. “I spent weeks going back through all the code. I thought I’d messed something up and would have to delay submitting. But then I showed my adviser, and he said, ‘Nope, your math is right.’” The math showed declines in three important testing categories, including matrix reasoning (abstract visual puzzles), letter and number series (pattern recognition), and verbal reasoning (language-based problem-solving). The first two, in which losses were deepest, measure what psychologists call fluid intelligence, or the power to adapt to new situations and think on the fly. The drops showed up across age, gender, and education level but were most dramatic among 18-to-22-year-olds and those with the least amount of schooling.
How Hustle Culture Got America Addicted to Work – Business Insider – in America, the long, steady march toward a more leisurely future came to an abrupt halt. Today, according to the international economic database Penn World Table, the German work year is an astonishing 380 hours shorter than ours — which means that Germans work almost 10 weeks less than we do every year.
Even stranger, Americans began to glamorize their lack of free time. As the boomer generation reshaped society in its own image, it brought its ’60s, countercultural ethos to the workplace — transforming the staid, conformist office into a vessel of self-expression. Work became the central means by which you undertook to live your best life, follow your passion, and change the world. As Goldman bankers and Google idealists alike began to toil through the nights and weekends that previous generations had fought so hard to secure for them, mental-health professionals bemoaned the rise of what became known as “hustle culture.” Working long hours was suddenly the ultimate status symbol, a peculiarly American form of humblebrag. In 2017, a clever marketing study found that if you told an American you worked long hours, they assumed you were rich. If you told an Italian the same thing, they assumed you were poor.
Waymo Has Come for the Kids in Los Angeles – The New York Times – “Here, it is not unusual for families to have multiple children attending different schools far from home. School buses, if you are deemed eligible, are limited to dropping off and picking up children at locations and times that are often unhelpful. The city bus, if there is somehow a direct route to school, comes with its own set of risks that can make parents uneasy.
Ms. Rivera, a psychiatric social worker, is stuck at work until 6 p.m. most days, while her husband, who installs and repairs glass, comes home even later.
The couple struggles to coordinate their jobs and their three children. They tried Uber, and Lyft, but found that those drivers tended to cancel after discovering their riders were minors. They turned to HopSkipDrive, a service geared toward students, but the drivers had to be scheduled in advance, and would leave if children were late.
Then, a few months ago, Ms. Rivera and Alexis did a test run with Waymo.
“It was the only option where I was like, ‘Oh my God, she can order a car, nobody’s in there, she can unlock it with her phone,’” Ms. Rivera, 42, said. “I know she’s going to be safe and she’s going to get home.” – interesting use case
Chinese luxury goes local | WARC – High-end Chinese brands are stealing a march on their Western rivals with homegrown labels that appeal to more discerning local consumers who are looking for luxury items that feel tailored to them. China’s $49bn luxury market is “changing fast”: ecommerce sales at jeweller Lapou Gold, for instance, have surged more than 1000% in the first three quarters of this year compared with two years ago. Songmont, a Chinese brand that claims to have ‘experiential’ designer bags, has grown its online sales 90% while Gucci online bag sales in China have fallen 50%, according to the Business Times. – This was inevitable when you had so many talented (and a number of mediocre) Chinese people being brought through the likes of Central St Martins.
Coca-Cola CMO Manolo Arroyo on WPP, AI and a new era for media | The Drum – Coca-Cola’s marketing ecosystem was sprawling and complex. The business was working with approximately 6,000 agency partners globally, while the majority of its multi-billion-dollar media budget was allocated to traditional channels. Arroyo wanted fewer partners, deeper integration and a shift towards digital-first execution at scale.
That ambition led to the consolidation of Coca-Cola’s global advertising account into WPP and the creation of Open X, a bespoke unit designed to manage the brand across markets and disciplines. Nine studios were established in key regions, housing a mix of Coca-Cola employees, WPP staff and specialist partners.
“It’s a marketing factory,” says Arroyo. “There are more than 2,000 employees of Coca-Cola and more than 2,000 employees of WPP […] and ultimately it’s enabled us to move from a company that in 2019 was investing close to 75% of our paid media on traditional TV, to a company that’s going to end up this year putting 70% of all our paid media on digital, particularly social and influencer led, marketing. For us, it’s our new TV.”
Outcry after French army chief’s ‘prepared to lose children’ warning | Le Monde – “We have all the knowledge, all the economic and demographic strength to deter the Moscow regime from trying its luck by going further,” said Mandon. “What we lack, and this is where you have a major role to play, is the strength of spirit to accept suffering in order to protect who we are.”
Paying tribute to French forces deployed worldwide, he added: “If our country falters because it is not prepared to accept – let’s be honest – to lose its children, to suffer economically because defense production will take precedence, then we are at risk.” – I don’t think that the west is ready or able to face Russia or China because of this. The war is lost before its fought
SOF, AI, and Changing Western Conceptions of War | Small Wars Journal by Arizona State University – Each generational shift in technology impacts military operations. Consequently, a shift in military training, command, and promotion structure should follow. Much of the conversation surrounding AI makes it seem like an unprecedented esoteric concept. While this is partly true, the same was said about steam engines during the Industrial Revolution. Simply put, AI is the next technological breakthrough and there will be more after it. As Clausewitz stated, the character of war changes, not the nature of war. A willingness to adapt while following strategic tenets will enable us to weather the storm and thrive in AI generation warfare. Failure to do so will only bring obsolescence while America’s adversaries gain global hegemonic status. Proper implementation of AI will result in faster decision making, more accurate intelligence, improved resource allocation, better spatial awareness, more effective messaging, and more impactful strategies. The key to reaching this level of success is SOF. SOF is uniquely equipped and trained to implement AI quickly and effectively, delivering results that can be scaled to the rest of the military.
A New Anonymous Phone Carrier Lets You Sign Up With Nothing but a Zip Code | WIRED – Phreeli, the phone carrier startup is designed to be the most privacy-focused cellular provider available to Americans. Phreeli, as in, “speak freely,” aims to give its user a different sort of privacy from the kind that can be had with end-to-end encrypted texting and calling tools like Signal or WhatsApp. Those apps hide the content of conversations, or even, in Signal’s case, metadata like the identities of who is talking to whom. Phreeli instead wants to offer actual anonymity. It can’t help government agencies or data brokers obtain users’ identifying information because it has almost none to share. The only piece of information the company records about its users when they sign up for a Phreeli phone number is, in fact, a mere ZIP code. That’s the minimum personal data Merrill has determined his company is legally required to keep about its customers for tax purposes.
Waking the Sleeping European Giant – by Matthew C. Klein | The Overshoot – “Europe” as a geopolitical entity does not exist. Instead of a strong and independent continent capable of securing the lives and freedoms of its citizens, Europe is divided into dozens of countries, all of which are too small individually to stand up to external threats. The problem is compounded by the mismatch between where the military resources can be found and where they are most needed. There is relatively little overlap between the places with the balance sheet capacity (mostly in the north), the places with the productive capacity (mostly in the center), the places with the largest populations of otherwise unoccupied fighting-age men (more in the south), and Europe’s front lines (largely, although not exclusively, in the east).
Bending Spoons raids the digital graveyard for paranormal returns | FT – businesses in the Bending Spoons stable: AOL, the dial-up internet service that had been most recently attached to Yahoo, and Evernote, the virtual scratch pad. – alongside Vimeo and Brightcove with Eventbrite due to join them
What I’m calling augmented retailing is a step change in technology in retailing from efficiency to effectiveness.
Retailing efficiency
Before we talk about augmented retailing, let’s go through efficiency which has been the focus for a long time. Depending on the way you want to look at this, you could look back centuries to the foundation of a technology multiplier: children.
Family retail businesses had family members working in their business from once they could understand and act on instructions. I have friends that started working in the family restaurant from 6-years old. I started working on the family farm (ineffectively) from a similar age.
Weighing scales were starting to become standardised by the middle-ages and giving short measures could see you punished.
The next innovation were payment related, such as currency and credit.
At the end of the 19th century thanks to industrialisation you started to see the origins of supermarkets as we now think about them. Sears Roebuck was the exemplar for mail order business, from which we now have Amazon and the countless e-tailers out there. Around about the same time the cash register was invented, which allowed cashiers to deal with more customers in a given amount of time. Cash registers then improved over the next century.
At the end of the 20th century we start to see e-tailing emerge. Accelerating mail order from being a niche to becoming a mainstream form of retail. Around the same time, you also saw cashier-less tills come in and other techniques to make shopping even more self-service.
Augmented retailing
Augmented retailing isn’t for the primary benefit of the retailer; but the customer. That’s more radical than it sounds as I write this down.
Look at other trends that marked change in retail and its about inconvenience to ‘create’ demand:
Apple retail product launches.
Abercrombe and Fitch / Hollister’s ‘club style’ door queues.
I have started to see innovations that are focused on the effectiveness of the consumer experience, rather than being orientated around retailing efficiency.
Mylowe virtual advisor
Lowe’s is a DIY superstore, rather similar to B&Q, Homebase or Toolstation in the UK. Like B&Q, Lowe’s has on-site experts to advise customers and help them select parts for projects.
Mylowe helps the experts by augmenting their expertise, providing a faster, better experience for Lowe’s customers.
A-eye
Indian snack manufacturer Britanniaworked with their agency VML India to aid blind consumers to shop independently. The app christened A-eye uses Google’s Vertex AI to recognise products on shelves and provide information about the product (quantity, ingredients, instructions etc.). Think about the personal confidence that this would bring to the user in their everyday life.
Albert Heijn provides cooking tips
Dutch supermarket brand Albert Heijn uses generative AI to help consumers by answering cooking questions. `Mijn AH assistent’ helps customers in their food shopping for food ingredients.
