The Before Time
Remember travel agents? In 1995: gatekeepers to the world of travel. By 2005: travelers booked directly. Intermediaries like these disappeared because they became unnecessary.
The Internet rewires who talks to whom.
Disintermediation is the word, part of the story of the Internet age. So far, it's happened in three distinct waves, each more profound than the last:
Web 1.0 (1995-2005): Amazon bypassed bookstores. Expedia bypassed travel agents. E*TRADE bypassed stockbrokers. Information gatekeepers lost their monopolies overnight.
Mobile Era (2007-2020): Uber eliminated taxi dispatchers. Airbnb sidestepped hotel chains. Social platforms connect creators directly with audiences. Smartphones rebuilt and built entire industries.
AI Era (2021-present): Now AI bypasses knowledge workers themselves. Legal research sans associates. Generated content. No code code. Technology removes intermediaries, disrupts industries, and creates new intermediaries.
Wither Middlepeople?
AI replaces connective tissue, and it collapses interface layers, insisting on direct connections:
- Need a recipe? ChatGPT delivers it instantly. Food blogs and search engines: bypassed.
- Writing code? AI generates it from natural language. Stack Overflow and documentation: done.
- Designing? On demand. Freelance marketplaces and designers: disintermediated.
AI eliminates entire support, curation, and organization ecosystems ā the infrastructure that previously connected creators to consumers.
Reintermediation
Folding layers creates both threats and opportunities.
The roles most vulnerable are those that primarily organize, curate, or lightly process information: the traditional "middleware" of the knowledge economy. Conversely, the roles most likely to thrive:
- Create the underlying systems and infrastructure (the platforms themselves)
- Provide expert guidance that AI can't replicate (deep specialist knowledge)
- Negotiate between human needs and AI capabilities (prompt engineering, system design)
- Curate experiences that can't be digitized (in-person events, physical products, unique services)
Successful AI implementation isn't about speed or efficiency but discernment. The most valuable AI systems won't be those that process the most data or generate the most content. Intelligent filters create value by making qualitative judgments that were previously reserved for humans.
Marketing (Making It) in an AI World
- Stop optimizing for search traffic and start designing for impression share and brand recognition within no-click environments.
- Build direct channels (email newsletters, SMS, private communities) that bypass search algorithms entirely.
- Create hyperlocal and experiential content that AI can't replicateāthings tied to specific locations, communities, and authentic business experiences.
- Position yourself either at the very beginning of the customer journey (the moment of need identification) or at the very end (the moment of solution implementation)āthe middle is being hollowed out.
The businesses that thrive in this third wave of disintermediation will fully redesign their operations around these collapsed interfaces. Think Amazon reinventing retail and Uber reimagining transportation.
Complexity: the Refuge of Small Minds
Artisanal marketers will find effective niches while automation dominates mass production with increasingly Rube Goldbergian AI tools. They'll differentiate on essential values of taste, discernment, and human connection.
15 Years a Client
If you ever wanted to get all the data on installed imagining devices, you probably called Radiology Data & Research. If you ever wondered how they got all the data, you'd be invoking a story that began at the end of the Aughts when Lambent built that data set from scratch. If you want to learn how it all happened, it's just a click away.
As Luck Would Have It...
AI and radiology have an odd history ever since 2016 when Geoffrey Hinton suggested that AI algorithms will replace radiologists. But that's not what happened.
Stripe Splaining Disintermediation
The payment unicorn might know a thing or two about parlaying opportunity into market dominance when encountering a sclerotic industry like online payments (RIP Braintree, Paypal, et al.) Here's what they have to say.
It's Probably Further Along Than You Think
Aneesh Raman, the chief economic opportunity officer at LinkedIn, notes that the unemployment rate for college grads has risen 30% since September 2022, compared with about 18% for all workers. Suggesting that entry-level knowledge work needs a radical shift in approach.
Meanwhile, Blood in the Machine's Brian Merchant minces no words when he states: "The AI jobs crisis has arrived. Itās here, right now. It just doesnāt look quite like many expected it to."
Unstocking Stock Art
We use a lot of stock at Lambent HQ, and stock selection is a major time suck. Wouldn't it be great if you could just describe what you need and have an AI tool produce a perfect replica of what's on your mind like a digital genie? It doesn't actually work that way, but that isn't stopping AI's economic impact on the stock photography industry. Kaptur.co muses that AI displacing 5-15% of demand would trigger up to $698 million in annual losses.
"With his co-authors Mr Kissinger has...used his vast experience and versatile mind to make a muscular contribution to one of the 21st centuryās most pressing debates." ā Economist review.
But seriously, The Age of AI: And Our Human Future is worth a read ā if frustratingly all over the place.
The "co-authors," Eric Schmidt and Daniel Huttenlocher, also bring experience and versatile minds to the party.
