The most interesting piece of news this morning comes from Microsoft’s Satya Nadella, who made a very public point of chastising his own team for saying the quiet part out loud.
That quiet part? In an internal memo leaked to 404 Media, a Microsoft VP said his team’s goal was to “make people addicted” to Microsoft’s new Scout tool, which is fashioned on Open Claw, the AI agent project that went viral early this year. Nadella quickly quashed such sentiments, releasing a memo stating “this is absolutely a non goal! If anything we are doing the exact opposite. We want to make sure AI empowers and adds real value to human endeavor and broad economic growth! We should make our teams clear about this.”
Microsoft then went into damage control mode, with corporate comms chief Frank Shaw piling on that Scout is for “helping people accomplish tasks more effectively—not encouraging dependency. Our goal isn’t more screen time. It’s more time back.”
I was standing at my bathroom sink, finishing up my daily ablutions, when a random thought popped into my head. It’s been a minute since I checked in on my favorite baseball team – ever since I moved East, they’ve been in a slump. Maybe they’re pulling back into contention this year? I still have a Google Home plugged in nearby. I’m intimately familiar with the Home’s limitations, but I asked anyway. Perhaps I was hoping one of my annual predictions (voice interfaces for the home) would magically come true.
It took me two weeks, 6000 words and nine posts, but I can finally round up my predictions for 2026 in one place. Here’s the complete list in one handy, blissfully shortened post. Thanks for reading, and once (and for good), I wish you a happy, healthy New Year.
The algorithmically induced sugar sludge that has dominated culture for more than ten years is failing, and 2026 will be the year most of us notice that trend. And the company that will be most impacted? Meta and its flagship Instagram app.
Anthropic has a cleaner structure, cap table, and set of relationships than OpenAI, which should allow it to be more nimble in its journey to becoming a public company. In addition, I sense Anthropic’s corporate culture would embrace the responsibilities and reporting requirements inherent to being a public company. Anthropic also has a better financial story to tell – it’s been focused from day one on the enterprise market, where it is a leader.
2026 will be a year of innovation when it comes to the cost of compute, as well in how much compute is actually needed to perform the magic we’ve come to expect from AI applications. It’s happened over and over again in this industry, and I think pricing the future based on the cost of the present is a losing bet.
For the past five or so years, tech giants have had to play defense when it comes to M&A and sweetheart partnerships – Meta was being sued over its acquisitions of Instagram and WhatsApp, Google over its consolidation of adtech and its domination of search distribution through deals with Apple and Samsung, among others. But in 2026, the governors are coming off.
By year’s end, we’ll have ambient AI in our homes, and it will actually work as expected. This in turn will shift what we expect as consumers of technology, in our cars, on our phones, and in the world around us. It won’t be a revolution, but when we look back at 2026 ten years from now, we’ll realize that this was the year “ambient intelligence” took off.
2026 will be the year that health takes center stage in the societal debate around AI. And OpenEvidence will be acquired by either OpenAI, Google, Apple, Microsoft or another advertising-driven big tech player (let’s not forget that Microsoft owns LinkedIn). In 2026, everyone will have a point of view on how health and AI interact.
This will be the year that AI moves beyond chatbots and into the fabric of everyday life in ways that will surprise and delight hundreds of millions of people, changing what they thought they could accomplish and reordering economic productivity along the way. That in turn will drive tectonic shifts in the business models underpinning most companies reliant on digital technology. There will be lots of magic this year. But there will also be plenty of carnage as previously unbreachable moats start to crumble, not only in business, but also in society at large.
This year, ideological battle lines will be drawn – you’re either in favor of the glorious future that AI promises, or you’re fighting the plutocrats seeking to cement their power through mechanisms of surveillance capitalism. The evolution of society with AI will be messy, ungoverned, and seemingly incomprehensible. But at its core lies an existential question: What happens to us when we build machines capable of fabricating reality, seemingly rendering us obsolete in the process?
In 2026, we’ll collectively decide if we trust generative AI, and any number of actors will prosper or fail based on that trust. Apple will leapfrog into the top echelon of AI companies, OpenAI will struggle, and large consumer brands will realize that in the context of AI sweeping through society, their most effective competitive advantage is the trust their brands evoke amongst consumers. 2026 will be the year that trust becomes the essential ingredient in business, culture, and society. And we’ll have the conjuration of generative AI to thank for the trend.
Cecco de Caravaggio The Conjurer (The Musician) c. 1600-1620
The modern English verb ‘to conjure’ is derived from the Latin conjurare, meaning ‘band together by an oath, conspire.’ Its roots con (‘with’) and jur (‘legal right or authority, law’) echo with questions central to our present day struggle with technology: Who do we trust to determine authority? Why do we believe in them?
Conjuring also evokes magic, sorcery, and wonder, essential elements of the tech industry mythos. My earliest pieces on the impact of generative AI leaned on the metaphor of magical “genies” doing our bidding in a relationship bound by loyalty and trust. Do those genies work for us, or are they the product of conjurers beyond our control? Do they demand faith, or instill it?
Today’s technology industry is defined by conjurers of the first order – they captivate us with stories of AI’s power, then justify their actions as inevitable outcomes of their narratives. Techno-optimism becomes religious text; beating China in the race to AGI becomes pretext for dismantling regulation, ignoring climate change, and redirecting capital from public good to private gain.
In 2026, all of us will be challenged to answer one simple question: Do we trust these conjurers?
In the United States, we’ve always been drawn to a good conjurer. Priests are conjurers, as are lawyers, directors, actors, musicians, politicians, writers, and lunatics. All are tellers of tales and weavers of possibility. We believe so deeply in the power of an individual to change the world for the better that we’ve mythologized the notion that anyone can succeed through hard work, perseverance, and honesty. We revere the “self made man.”
But no class of conjurer has captured our imagination quite like the modern entrepreneur. These spell casters weave words into facts, imagination into capital, and capital into action. What begins as a story becomes a company made real by a band of conspirators. Apple, Google, Meta, Amazon, Tesla, OpenAI – these are world-changing companies birthed of incantations, faith, and capital. We celebrate them as proof of our shared belief in the American dream.
—
2026 will mark the year that conjurers are no longer solely human. For the first time, we confront a set of technologies capable of conjuration independent of human constraint or comprehension. We’ve built story machines that can entertain, inform, advise and manipulate us – machines designed to conjure anything we care to consume.
Do we have any idea how to handle such a beast?
As Stewart Brand famously declared decades ago as he contemplated the impact of technology on society, “we are as gods, and might as well get good at it.” 2026 will be the year we find out how good we’ve gotten.
I’m not sure how this story will play out, but I am sure of one thing: its ending will be determined by trust. Trust is sometimes a slippery concept, but I like this definition: “The willingness of a party to be vulnerable to the action of another party based on the expectation that the other will perform a particular action important to the trustor, irrespective of the ability to monitor or control that other party.”
In 2026, we’ll collectively decide if we trust generative AI, and any number of actors will prosper or fail based on that trust. And since this is supposed to be a predictions post, I’ll give you a few specific examples of what I mean:
Leveraging its market position as an advocate for individual privacy and safety, Apple will leapfrog into the top echelon of AI companies by cleverly productizing (and commoditizing) generative AI across its walled garden ecosystem. Its core message will be “Trust Us – We’re Apple.”
OpenAI will struggle throughout the year. It will try to do too many things and be constantly distracted by its governance structure, to be sure, but as a young company driven by a “move fast and break things” mentality, it will lose the confidence of consumers, many of whom will opt for brands they feel they can trust.
Large consumer brands – think P&G, Nestle, Coca Cola, JP Morgan Chase – will realize that in the context of AI sweeping through society, their most effective competitive advantage is the trust their brands evoke amongst consumers. Expect many of them to launch marketing campaigns positioning themselves as trusted partners, not just as purveyors of superior consumer products.
In short, 2026 will be the year that trust becomes the essential ingredient in business, culture, and society. And we’ll have the conjuration of generative AI to thank for the trend.
—
This is the ninth and final in a series of post I’ll be doing on predictions for 2026. The first eight are here, here, here, here, here, here, here and here. My next post will be a short roundup, like I usually do.
Ten years ago a new and promising technology burst into our homes – the smart speaker. Like many tech-forward families, our household went all in. We got two Alexa speakers and two Google Homes, plugged them in, and they became fixtures in our kitchen and bedrooms for years.
Problem is, we kind of hate them now. At first they were cool – it was novel to talk to a device and have it actually work, at least for simple tasks like “what’s the weather today” or “play Vampire Weekend.” But we quickly grew disaffected with our new purchases, because more often than not, they failed when presented with even moderately complicated queries like “what time is the Giants game tonight” or “what’s on my grocery list.” In short, the first generation of smart home speakers were limited by a rigid approach to “intelligence” that didn’t scale. Only one sad, bedraggled Google Home remains in service in our kitchen, serving as a glorified clock radio (that’s it in the picture above). And it’s not doing Google any favors in the branding department, because whenever we ask it anything even slightly complicated, it fails, earning a string of expletives in the process*.
But all that is poised to change in 2026. We’re now all in the habit of having nuanced, complicated, and satisfying conversations with AI chatbots, and in the past year, those conversations have increasingly taken place on our phones, using a voice interface. Home speakers have been an obstinate exception to this new habit, but this year, the three major smart home players – Amazon, Google, and Apple – will finally integrate conversational generative AI into the next generation of their devices. If they work properly, we’ll finally get the voice-driven interface revolution I’ve been excited about for decades.
My prediction is this: By year’s end, we’ll have ambient AI in our homes, and it will actually work as expected. This in turn will shift what we expect as consumers of technology, in our cars, on our phones, and in the world around us. It won’t be a revolution, but when we look back at 2026 ten years from now, we’ll realize that this was the year “ambient intelligence” took off.
*I often imagine the poor QA engineers at Google or Amazon who have to listen to the snippets of audio the devices capture after their devices deliver reliably crappy results. I bet it’s gold.
—
This is the fifth in a series of post I’ll be doing on predictions for 2026. The first four are here, here, here and here. When I get to #1, I’ll post a roundup like I usually do.
All year long I monitor my annual predictions, taking note when events either make me a fool or a sage. 2025 marked perhaps the most unpredictable and frustrating year of them all – and that’s not nothing, given I started prognosticating in 2003. But then again, I did expect an odd one – from my 2025 post: “This isn’t going to be a normal year. 2025 will be strange, frenetic, and full of surprises.”
I titled my post “Tech Takes the Power Position.” While I didn’t make that sentiment one of my specifically numbered predictions, it did provide the context for how I was thinking about the year ahead. “We’re not accustomed to the tech industry having this much raw power. The finance industry? Sure…But this year, for the first time ever, Big Tech has leap-frogged finance in the pantheon of political influence…the subset of Big Tech bros who’ve bought their way into the Oval are evangelists for an untested and downright strange brand of magical thinking best summed up as “techno optimism.” …for better or for worse, 2025 is going to be the year when the loudest voices in the room are all adherents of the Great Man Theory, and they all happen to have direct access to the Oval Office.”
I certainly think that sentiment has proven out over the past 12 months – the tech industry is an indisputable driver of the tragicomedy currently playing out across American society.
But what of my specific predictions? Let’s roll the tape:
TikTok will continue to operate in the US. Check. Not only did TikTok continue to operate, the administration recently announced a sweetheart deal that addresses none of the original issues raised (by Trump!) in the first place (save making various cronies rich on the arbitrage). I’ve never been a fan of the TikTok ban, nor of TikTok itself, but I certainly got this one right.
There will be no meaningful regulation of Big Tech. Check again. Perhaps this was too easy to predict, and I should have gone further: There will be meaningful deregulation of Big Tech. At the state level, more than 30 AI related regulations passed, but that was greeted with an executive order attempting to reinterpret the Constitution so as to mollify Trump’s tech bro donors.
2025 will not be the year AI agents take off. Check. Three in a row! If you are merrily employing agents to do your bidding across the Internet, please let me know. So far, consumer agents remain a pipe dream, and enterprise agents have been slow to move beyond the kind of automations most companies already have hard coded. “No one knows what the hell an AI agent is” declared Techcrunch last spring. “Agentic commerce is still a collective hallucination,” proclaimed industry trade Marketecture this Fall. I have written about this story all year long and will keep focusing on it next year, so if you want more, just stay tuned!
2025 will be the year Gen AI gets boring – and better. Check. I think this also proved true. The term “boring” is certainly disputable – but my point was that it would be a year of evolution, not major leaps forward, and that’s certainly been the case. Products got better, usage kept going up and to the right, and, as I wrote: “2025 is the year Gen AI is put to work doing boring, useful things for us.”
Prompt data becomes the new gold standard.Push. This one is harder to support. First a bit of explanation – in my prediction, I wrote “prompt data provides one of the richest signals of what people want, need, and plan to purchase, and that is simply too valuable to not be leveraged by marketers (and Big Tech).” But did that happen this year? Certainly by Google and OpenAI, to both improve their AI products and develop their advertising systems. And I’ve met with dozens of interesting startups that are working in the space in one way or another. But did prompt data become a fungible class of data like, say, retail or financial data? Not really. I’m going to give myself a push here. I think I was right, but can’t prove it yet.
Retail media will consolidate. Push. There were too many deals to list over the past 12 months related to the retail media network world. But here are a few: DoorDash & Symbiosys, Pinterest & TVScientific, Instacart & OpenAI, Kroger & Omnicom; oh, right: Omnicom & IPG, which was driven in part by the data-driven nature of RMNs. But then again, we saw a bunch of new RMNs either launch or expand significantly, including Home Depot (OrangeApron), Ace Hardware (RedVest – see a trend here?), Best Buy Marketplace, Kroger Precision Marketing, and “shoppable TV” from Walmart’s Vizio integration. So perhaps true consolidation in this space is still a year or two off…I’ll give myself half credit on this one for now.
Apple will be in open warfare with OpenAI. Whiff. I was sure Apple’s 2024 OpenAI deal was provisional, and that Apple would end up selling its “consumer surfaces” to the highest bidder (IE Google). But that didn’t happen in 2025 – likely because both parties were waiting for Google’s antitrust remedies related to their search megadeal. Now that the suit has been resolved (quite favorably for both), I’d look for more Google-Apple news in 2026. And while I’m quite sure Apple is no fan of OpenAI, it keeps those feelings private. “Open warfare” never really broke out. My first big miss.
A Trump/Musk fallout? No. A burnout? Yes. Push. Well, this one is tricky. Was there a fallout? Yes. But did it settle down into a negotiated truce (a “burnout”)? Again, yes. Here’s what I wrote: “By year’s end, Musk and Trump will have tired of each other, preferring to do business with each other through proxies…both Musk and Trump are smart enough to realize they need each other – so they’ll avoid an all out press battle.” For the most part, that is how the story played out, but it did get ugly there back in the spring. I think I got this one mostly right, but missed what should have been obvious: When two billionaire narcissists split, they’ll always be a spectacle.
Google gets a new CEO. Major Whiff. In fact, just the opposite happened. When Bloomberg writes a piece titled “Sundar Pichai Is Google’s Wartime CEO After All,” you know you missed the mark, even if it seemed like you were correct at the start of the year. I’ll take the L here, and tip my hat to Sundar’s extraordinary comeback.
Health at the center. Check. This was a pretty broad prediction, but my main point was this: “Next to tech politics, the healthcare industry will be the most interesting story of 2025.” I stand by that statement, even if it’s entirely subjective. I had a front row seat to that story all year long through DOC, and the pace is only increasing. From policy (RFK’s vaccine panels, Medicare and ObamaCare at the center of a government shutdown) to breakthroughs in AI and cancer, women’s health, and many other fields, 2025 was a huge year for health. I expect that to only continue.
Crypto goes sideways.Check. With this prediction, I was guessing that the crypto craze would push prices up and up, but as with all pump and dump schemes, the market would end up finding the floor by year’s end. And that’s pretty much what happened. Bitcoin, which serves as a proxy for the market overall, started the year at around $92,000. It climbed to nearly $125,000 by mid year, but has been languishing back at $88,000 or so as of this writing. But wasn’t it a fun ride?!
Ok, checking my self-graded homework – that’s 7 of 12 right, 2 of 13 wrong, and 3 pushes. Not bad! Look for my 2026 Predictions in a week or so, and have yourself a merry holiday season!
Finance has always leveraged technology – at Wired in the early 1990s, we were fond of saying that technology’s twin engines of innovation were money and sex – but the most interesting story was always money. Care to understand the future of internet infrastructure? Bone up on how hedge funds optimize network latency. Want to peer into the future of online consumer services back when the Web was a glint in Marc Andreessen’s eye? Start with online banking.
But in the past few years, the bit has flipped. Technology is no longer a mere enabler of finance. Thanks in large part to society’s rapturous embrace of AI, technology has absorbed the finance industry, adopting its amorality and its rapacious appetite for risk. Capital has become tech’s hand maiden, financing an endless stream of audacious bets: Half a trillion for AI server farms? Let’s do it twice! A $2 billion seed round for a company with no product? Probably undervalued! Abandoning your values to access autocratic blood money? Everyone else is doing it, why not us!?
Hubris.
Technologists have always suffered from hubris. For a few decades, that was OK – you could poke fun at breathless coverage of digital hippies or the dog walking app that raised $300 million only to file Chapter 11 a few years later. And you could roll your eyes at a real estate con dressed up as a world-changing economic force for good. But today’s AI industry? Hubris seems too light a word.
The word hubris comes from the Greek for “a serious offense against the gods and the natural order.” Its modern usage evokes downfall – “extreme or excessive pride or dangerous overconfidence and complacency, often in combination with (or synonymous with) arrogance.”
Keep that definition in mind while we review the news over the past week or so:
It’s AI week in Washington DC, a town famously driven by power, money, and hubris. And OpenAI CEO Sam Altman has a message for Our Dear Leader: Bad guys with bad AI are willing to spend “an enormous amount of money” to overtake our beloved democracy. The only protection we have? Companies like OpenAI. So for the love of (God? Money? Power?), leave us alone to do (God’s? Money’s? Power’s?) work! Everything we hold dear depends on it, don’t you see? Copy that, the administration announced today. We got you.
OpenAI’s products are “the greatest source of empowerment for all” ever created, according to the company’s incoming CEO of Applications Fidji Simo. In an essay explaining her move from Instacart (boring, old school!), Simo recounts the tech industry’s de facto religious canon: AI will conquer every single problem known to humanity, including healthcare, economic opportunity, creative fulfillment, and …. time itself! Sounds great! Never mind that AI doesn’t really work as advertised. It will!
As Axios puts it this week: “AI makers are getting everything they have ever asked for or could possibly want.” Buried in that story is a line worth repeating: AI makers liken the process of training models to raising children. If that’s true, the technology is growing up as a fabulously rich kid in a wildly permissive household. Remind you of anyone?
In a leaked memo, former industry good guy and Anthropic CEO Dario Amoedi explains why he’s now open to accepting “blood money” from autocratic sources. In short: Everyone else is already doing it. “There is a truly giant amount of capital in the Middle East, easily $100B or more,” he writes, clearly earmarking that $100 billion for Anthropic alone. “If we want to stay on the frontier, we gain a very large benefit from having access to this capital. Without it, it is substantially harder to stay on the frontier.”
Data.
Where, exactly, is this frontier bringing us? As far as I can tell, it’s (a lot) more of the same. If anything about how the technology industry has behaved over the past 20 years gives you pause, well, it’s about to get several orders of magnitude worse. The most powerful industry in human history now has access to unlimited resources, unlimited political power, and an unlimited remit to “change the world.” But the business model it’s employing to get there? Same as it ever was.
Have you read OpenAI’s privacy policies? I have. At its core, it’s the same bargain we already have with the tech industry. Scholar Shoshana Zuboff calls it surveillance capitalism. Author Cory Doctorow calls it “enshittification.” It follows a simple formula:
We give you a service. You give us your data. We use that data to lock you in. Then we rent seek for as long as we possibly can.
We’re about to be nostalgic for the time when “our data” meant simply our name, our search and browsing history, our location, and what we bought on Amazon. As I’ve said over and over, the database of our intentions keeps expanding, and it now includes our increasingly unhinged conversations with AI agents.
How large is that last category? Axios reports this week that 500 million weekly active ChatGPT users “send more than 2.5 billion prompts each day globally.” This figure is already too large to comprehend, and it’s growing faster than any product in the history of digital technology. We are literally flooding the AI industry with our deepest fears, hopes, wants and dreams. What do we expect they’ll do with them?
If you don’t think OpenAI (and the rest of the AI industrial complex) plans on using that data to lock you into using its services, not to mention delivering increasingly intrusive advertising, well, as I said last week, you’re asleep.
Faust.
Perhaps, as many in the tech industry argue, we truly are at an inflection point in human history, and we’re building a new society based on entirely novel cultural norms and values that will usher in a time of prosperity, equality, and justice. Perhaps the techno-utopia we dreamt of in the early years of the Internet is just around the corner. Perhaps Sam Altman, Marc Andreessen, Peter Theil, JD Vance, and Elon Musk will take us there.
Perhaps. But our society’s bargain with the men behind AI remind me of the centuries old tale of Christopher Marlowe’s Doctor Faustus, later retold as an epic poem by Goethe, perhaps Germany’s most revered literary figure. In both versions, Doctor Faustus is bored with humanity’s limitations and seeks the omnipotence of unlimited knowledge and endless worldly pleasures. He strikes a deal with the devil, and for a period of 24 years, is granted his wish. In Marlowe’s telling of the story, Faustus must face the consequences of his bargain. At the end of his life, he is summarily dragged off to Hell. But in Goethe’s version, Faustus gets a mulligan. He repents and is granted entry into Heaven.
I get the sense that Goethe’s version of the tale underpins our current relationship with AI. No need to ask for permission – it’s already been granted by the platforms that built the tech industry. We’ll take every word every written, every image every created, and every last byte of data spun in real time from the minds and actions of billions. And if things go wrong later – not that they ever will! – we’ll just ask for forgiveness, and Heaven awaits no matter what. It’ll all work out! We’ve got money to spend, data to exploit, and worlds to conquer! Let’s gooooooooooo!!!!!!!
According to scholar Carlota Perez, one of tech’s most revered theorists, society regularly goes through technology-driven “revolutions.” These structural cycles can take fifty years or more, and are defined by core technologies which shape life as we know it. Her list of previous cycles include the Industrial Revolution; The Age of Steam and Railways; The Age of Steel, Electricity and Heavy Engineering; and The Age of Oil, the Automobile, and Mass Production.*
Back in the early 2000s, Perez has identified the Internet (more formally, ICT, or “information communications technologies”) as the dominant technological force driving our current age. Perez’s framing has been a favorite of pundits ever since – and has played a central role in the debate as to whether a much-hyped “Next Big Thing” – crypto, the metaverse, quantum computing – is merely a feature of an ongoing revolution, or the starting gun to an entirely new age.
Given the extraordinarily high volume of hype and investment in AI over the past few years, it’s fair to ask: Does Perez’s framing indicate we are embarking on a new age that supplants the Internet – are we entering “The Age of AI?”
It’s an important question, because as Perez and many others have noted, how we invest in, regulate, and incorporate new technologies differs dramatically depending on the answer.
For her part, Perez does not believe AI represents a new age separate from our ongoing Internet age, and I find myself agreeing with her. Unfortunately, technology’s most vocal leaders find it to their benefit to argue the opposite case – and that, I fear, will lead to poor policy decisions that will retard social flourishing and lead to a dark period of history in which autocrats prosper and democracies decline.
I know, that’s a depressing thought. Lighten up, Battelle! But everywhere you look, Extremely Rich Dudes are using AI as a enabler and/or a stalking horse for sweeping policy decisions: Musk to create a national database of undesirables, Altman (and nearly all his peers) to deregulate big tech, and Schmidt (among many others) to build China up as The Enemy Over Which We Must Prevail Before It’s Too Late.
So again, does AI represent a new era, or is it simply an important feature of an ongoing revolution? To explore that question, we need to dive a bit deeper into Perez’s framework. If we’re going to make the right decisions about what to do in the world of tech policy, we need to at least consider how history might judge our current moment.
***
Perez breaks tech’s phase shifts into four distinct phases, ranging from “gestation” – when new technologies are in an early stage, to “constriction” – when those same technologies are ubiquitous, mature, and innovation slows considerably. But she further points out that inside this four-phase cycle, there are many more identifiable moments, including Irruption – when the technological trend breaks out and begins to impact society – and a “Turning Point” – when the trend shifts from confusing and disruptive to coherent and integrated:
There are two distinct periods in each revolution – one before the Turning Point, and one after. “Installation” is where new technologies emerge, are supported by a frenzy of capital, and become threats to status quo structures in business and society. “Deployment” is where these same technologies become the status quo.
Installation is the time when the new technologies irrupt in a maturing economy and advance like a bulldozer disrupting the established fabric and articulating new industrial networks, setting up new infrastructures and spreading new and superior ways of doing things. At the beginning of that period, the revolution is a small fact and a big promise; at the end, the new paradigm is a significant force, having overcome the resistance of the old paradigm and being ready to serve as propeller of widespread growth.
The second half is the deployment period, when the fabric of the whole economy is rewoven and reshaped by the modernizing power of the triumphant paradigm, which then becomes normal best practice, enabling the full unfolding of its wealth generating potential.
As I’ve already noted, most observers of the technology industry – including myself, for the past ten or so years – have presumed that we are well into the Deployment phase of the Internet revolution. It’s hard to argue otherwise – after all, Big Tech dominates Wall St., culture, and politics now. It’s consolidation of power seems complete.
This is a seductive view, and if correct, it leads to an even more seductive conclusion. If the Internet revolution has matured to the point of saturation, then AI must represent the next technological revolution, one that will do to the Internet what the Industrial Revolution did to farming. Right?
Maybe not. Given Perez’s framework, the introduction of generative AI less than three years ago can be seen as simply one more step in the Internet’s march toward deployment. In fact, it may well mark a “Turning Point” – a technology that bridges the first few decades of Internet disruption with a future period where humanity can reap the Internet’s benefits.
The case for AI as a Turning Point only holds, however, if AI in fact delivers on its ambitious promises of mastery over data. If the first phase of the ICT/Internet revolution was leveraging computing to turn the world to data, the second phase will have to deliver mastery of that data to all of society, and not just to a strata of oligarchic elites. AI may be the struck match, but the Internet is the fire itself. And we’d be wise to not outsource the power of that fire to a cadre of evangelists bent on convincing us that they alone can save us from the flames.
I’ll ask you kindly to get the fuck off my lawn now.
I’ve been pondering something for a while now, but have held off “thinking out loud” about it because I was worried I might sound like a guy yelling at the kids to get off his lawn. But f*ck it, this is my site, and I think it’s time to air this one out: Technology isn’t delivering on the magic anymore. Instead, it feels like a burden, or worse.
For decades, digital technology delivered magical moments with a regularity that inspired evangelical devotion. For me, the very first of these moments came while using a Macintosh in 1984. Worlds opened up as that cursor tracked my hand’s manipulation of the mouse. Apple’s graphical user interface – later mimicked by Microsoft – was astonishing, captivating, and open ended. I was a kid in college, but I knew culture, business, and society would never be the same once entrepreneurs, hackers, and dreamers starting building on Apple’s innovations.
And build they did. From that point onward, the magic continued, sometimes in massive leaps (like the Mac itself), sometimes in smaller ways that built upon those leaps. One such example was my first interaction with desktop publishing software, in 1987. Learning how to work with that software turned me from college hack to professional media creator, and I never looked back. What I didn’t imagine was what the tech industry would do to the media world I was so eager to be a part of – but that disruption was still decades away.
My first login at the WELL, an early online service, was another such milestone of the late 1980s. Staying up late talking about random Grateful Dead shows with people based all over the world made me realize how powerful “social networks” were going to be. I almost felt like I was getting away with something – like I had a secret shared only by a few thousand others. I imagined a day when services like the WELL would sport robust graphical user interfaces and unlimited bandwidth. But again, what I didn’t imagine was the centralized, algorithmically driven models of Facebook, Instagram, and TikTok. The WELL – a command driven, text-based service – was social media at its best, a funny thing to reflect on given all that’s come since.
Then there was the web. Those of you who came of age after the web became commonplace can’t imagine how mind-bending it was to be one of the first people on earth to load a web page and jump from link to link. In 1993 I downloaded an early version of the Mosaic browser and began a lifelong journey. But back then, most of the sites were earnest, information rich, amateur, and just plain fun. Yes, the early web got super crowded, super fast, but that’s when another piece of magic broke through – Google.
Google in 1998 was a magic machine equivalent to the Mac in 1984. You put in your question and *boom* – nine times out of ten you ended up where you wanted to be. It was captivating, liberating, and incredibly useful. It also sparked the rise of the independent web – an early (and mostly lost) version of the internet that was driven, for a few years, by a relatively pure signal called backlinks. Which leads to the tech’s next magic moment….
Blogs. I know, just typing that word kind of makes me cringe, because I’ve always hated the term. But from about 2001 to 2008, weblogs formed a truly magical ecosystem where respectful debate and true conversation took place. Blogs formed the spine of the Internet itself, giving shape and substance to Google’s growing index (and dominance).
Tech offered some non-internet related magic as well. If you’ve never used a Tivo, then you unfortunately missed out on television’s most singular moment of product-market fit. From 1999 until the mid 2000s, Tivo worked like a blanket digital recording device that swallowed all of cable television and re-ordered it in a way that you, the consumer, controlled completely. You could record whatever you wanted, time-shift your viewing habits, and skip the ads with a few clicks of your thumb. It completely changed what I expected from television – and since Tivo’s demise, my TV viewing experience has only gotten worse, despite moving almost entirely to streaming. The reasons for this are myriad, but in essence they come down to this: The television industry didn’t want Tivo to dis-intermediate its relationship with consumers. Funny what happened instead: Apple, Google, Amazon and Netflix stepped in and did it anyways – and far less well.
Tivo makes me think of another brief moment of tech-driven magic: Early Spotify. It might be hard to imagine now, but in 2008, when Spotify launched, most music fans had large collections of compact discs – CDs – and most computers had CD-ROM drives. That meant you could take your entire music collection and rip it to your computer, then upload it to Spotify, which offered a robust interface for managing and playing your music. I used to buy new music on Amazon, tear open the box, toss the CD into my Mac, rip it to Spotify, and listen to my new tunes over and over. It was just like buying a new album, but the digital interface gave me new superpowers like playing music on my computer, searching through music based on keywords and playlists, creating “mixtapes” on the fly, and much more.
There are many, many more such moments – Skype, the first Kindles, early usenet, hell, even early Facebook, early Snap, and the first iPhone – but I’m going to stop now and think out loud a bit about all of this. What all these moments of tech magic have in common is the power and control they gave to the average person. Early tech gave us agency over parts of our lives that were previously hobbled by physics and economics. The Mac gave computers a physical interface. Desktop publishing allowed us to leap over the physical and economic limitations of the printing press. Early online services, then the web itself erased boundaries of geography, information access, and culture. Google, for a brief and wonderful period, helped it all make sense. Services like Tivo and Spotify gave us agency and elasticity over media.
But..I no longer feel like Google, or the web, or television, or Spotify are in anyway magical. In fact, they all kind of suck now. Why? Yes yes, the ad model and surveillance capitalism, but honestly, I think it comes down to our society’s cardinal sin of favoring convenience over agency. Turns out it’s just easier to let Spotify or Netflix or Instagram feed us media that, while perhaps not our first choice, seems consistent with things we’d likely choose if we actually decided to make a choice in the first place. The magic that we felt when tech was young has been replaced by the dark art of the hidden and inscrutable algorithms busily feeding us stuff we never would have thought of engaging with otherwise.
I fear this model – where our own agency has been sacrificed at the altar of convenience – will serve us poorly should artificial intelligence become the next framework for how we engage with computing going forward. And there’s no doubt that Nearly Everyone In Tech wants generative AI to be The Next Big Thing, proof that tech can, in fact, bring back the ol’ magic of yore. I’ll expand on this idea more in future posts, but in short: If we don’t take our agency back when it comes to “the agentic web,” I fear this entire AI revolution will come to naught. And we’ll have wasted billions – and a helluva a lot of our collective time – in the pursuit of it.
(This is a preview of a piece I’m working on for Signal360, to be published next week.)
“The US litigates, the EU legislates.” That’s what one confidential source told me when I asked about theDigital Services Act and theDigital Markets Act, the European Union’s twin set of Internet regulations coming into force this year. And indeed, even as the United States government continues anendless parade of lawsuits aimed at big tech, the EU has legislated its way to the front of the line when it comes to impacting how the largest and most powerful companies in technology do business. It may be tempting to dismiss both theDSA and the DMA as limited to only Europe, and impacting only Big Tech, but that would be a mistake. It’s still very early – much of the laws’ impact has yet to play out – but there’s no doubt the new legislation will drive deep changes to markets around the world. And even if you aren’t a digital platform, your own business practices may well be in for meaningful change.
So what do the DSA and DMA do? Both pieces of legislation target “big tech” – most of the targeted companies are in the US – and require them to enact novel forms of accountability for how both consumers and businesses interact with digital services. The DSA, which came into force in August, targets “very large online platforms” with more than 45 million users in the EU – sites like Meta’s Facebook and Instagram, Apple’s App Store, Bing, Google Search, Microsoft’s Bing and LinkedIn, Snap, Twitter, and TikTok. The DSA’s goal is to define online services’ responsibilities related to content moderation, including new rules around use of algorithms and data, user choice, annual audits for compliance, and advertising to minors. In short, the DSA seeks to make the Internet a safer and more transparent place to shop, do business, and be entertained.
The DMA came into effect this past May, and focuses on “unfair” and anti-competitive behavior by the largest companies on the web – what the EU calls “gatekeepers.” These include Amazon, Apple, Google, Microsoft and Meta. Many of the DMA’s provisions address the same behavior that has prompted various US agencies and states to sue – Apple’s refusal to allow competitive app stores, for example, or Amazon and Google’s alleged practice of favoring their own products and services over those of competitors. Under the DMA, Apple and Google can no longer force app makers to use their app stores, for example. Platforms are required to share data with their customers, obtain explicit opt-in consent to use data across services, design systems that are interoperable, and are barred from using their data to gain competitive advantage.
That means search results for just about every major platform in the EU – whether it be Google, Amazon, Microsoft, and even Meta, will be changing soon. “We are still at the level of philosophy,” as to how those changes might look, said one tech company insider who asked to remain anonymous. “The laws are not yet being enforced.” But what’s certain, he continued, was that “we need to provide new choices and spaces for consumers and competitors.”
The two pieces of legislation are dense with theory and explication, but the core intent is clear. As the DMA puts it, large online platforms have led to “serious imbalances in bargaining power and, consequently, to unfair practices and conditions for business users, as well as for end users of core platform services provided by gatekeepers, to the detriment of prices, quality, fair competition, choice and innovation in the digital sector.”
In short, the EU isn’t messing around. Importantly, the new laws require that platforms be in compliance by early next year, and they must continue to prove compliance on an annual basis. And the laws create full time regulators responsible for enforcement and fines, which are steep – up to 10 to even 20 percent of a company’s EU turnover. That means tech companies can’t see fines as a cost of doing business. Net net: a lot is going to change over the next two quarters.
So how might the new laws change business for non-tech companies, both inside and outside the EU? The most direct impact will be for marketers – if you’re targeting children under 18, you’ll lose access to that personalized data on all major platforms. You may also lose cross-service data – between Google Maps and YouTube, for example – if companies like Google fail to get explicit opt in from their customers.
In addition, the DSA requires that platforms create up-to-date repositories of data on advertising purchases, exposing the strategies and investment levels of every advertiser on the platform. “The DSA requires us to give almost live information about the amount of money being spent on ads,” said the tech company insider. “I think that will be interesting for marketers.” Indeed it will be – that kind of information was previously considered top secret, and will certainly re-shuffle go-to-market strategies for most CMOs.
If you’re a vendor on Amazon, for another example, you’ll suddenly find yourself free to compete on a more level playing field. The threat of Amazon using its data to undercut you on pricing, or to create and market a generic version of your branded product, is now gone in Europe. “Amazon is not going to be able to use data from other shampoo makers to compete against P&G,” said Fiona M. Scott Morton, the Theodore Nierenberg Professor of Economics at the Yale School of Management. “And P&G will have access to tools to access, evaluate and track advertising that they place on site,” allowing the company to evaluate how their investments are performing relative to historic and industry norms.
But the true impact of the DSA and DMA may be in how the overall business ecosystem adapts over time, and this is where it pays to imagine a few out-of-the-box scenarios. Imagine, for example, that the DSA and DMA work well in the EU, and companies of all sizes begin to demand similar types of affordances in the other large markets. Might legislators adopt similar regulations once they are pressured by the likes of P&G, Walmart, or Nestle? It’s possible, says Daphne Keller, Director of the Program on Platform Regulation at the Stanford Cyber Policy Center. More likely than not, “whatever big picture changes they make to appease EU regulators will end up being done globally.”
Beyond the potential for the DMA and DSA to become de facto standards outside the EU, there are also subtle, insistent market pressures to consider. Prior to the regulations taking force, no company – whether startup or large platform – would have ever attempted to create an app store that competed with Apple. But now, “Facebook could just create an advertising program that directly installs apps on a person’s phone, bypassing Apple altogether,” points out the tech company insider. Apps installed in this fashion would not pay Apple’s 30 percent tax on app revenue – a powerful new incentive for entrepreneurs to innovate.
In short, don’t sleep on the EU’s new DSA and DMA regulations – they will not only change how consumers interact with large platforms, they may also end up changing the rules of business on the Internet for good.
The most interesting piece of news this morning comes from Microsoft’s Satya Nadella, who made a very public point of chastising his own team for saying the quiet part out loud.
Capture.
There are two distinct periods in each revolution – one before the Turning Point, and one after. “Installation” is where new technologies emerge, are supported by a frenzy of capital, and become threats to status quo structures in business and society. “Deployment” is where these same technologies become the status quo.
