- The Angle
- Posts
- The bottleneck cascade
The bottleneck cascade
David Peterson
In April 2024, Mark Zuckerberg told Dwarkesh Patel that Meta’s biggest constraint in developing AI wasn’t GPUs. It was power. Even with unlimited capital, the lead time on new energy generation was years.
Tyler Cowen made a similar point. When asked why the AI-takeoff may be slower than people expect, he argued that the limiting factors aren’t technical, but human (i.e. cultural and regulatory). For example, even if AI increases the number of high potential drug targets by 10x, the FDA still needs to approve them.
What these examples reveal is a pattern I call the Bottleneck Cascade. Every technological breakthrough makes one input abundant, and in doing so it pushes scarcity somewhere else in the system.
In other words, progress doesn’t eliminate constraints. It moves them.
History offers countless examples of cascades in action. Here’s a particularly relevant one: in the 19th century, factories had to be built next to rivers to harness water wheels (or else rely on steam engines). The arrival of centralized electricity generation and distribution blew the bottleneck open. Suddenly, you could pipe power into any building in any city, and factories no longer needed to be clustered by rivers.
Electricity didn’t transform manufacturing overnight, however, because cheap electricity created new bottlenecks: the need for transmission infrastructure, new machinery, and above all, new factory designs which took advantage of this strange, new power source.
That’s the pattern consistently seen with general purpose technologies. When a GPT first arrives, productivity often falls. Its real-world utility depends on complementary innovations. In this example, machines with electric motors needed to be invented and “electricity-native” factories needed to be designed before electricity’s full impact could be felt. That took decades.
And indeed, this is the exact dynamic we’re living through today. AI is the most powerful technology of our lifetimes, but its trajectory will be defined not by raw capability, but by the sequence of bottlenecks it collides with. Two years ago, GPUs were the scarcest resource. As supply ramped, the bottleneck shifted to power. What else is holding AI back? Perhaps talent (how many people in the world can reliably train frontier models…potentially fewer than 1000)? Maybe high-bandwidth memory? What about actual real-world implementation of AI (see the much-discussed MIT report on the failure of enterprise generative AI pilots)?
The cascade continues.
Seeing the world through this lens is useful because it provides a hint as to where value might accrue. Track enough cascades and you’ll realize that the company that breaks a bottleneck is not the one that captures the value. Containers revolutionized shipping, but the container makers didn’t become giants. The winners were the abundance-native firms like Walmart. Walmart won not by making shipping cheaper, but by mastering the scarce capabilities that mattered once it was: logistics, purchasing power, and scale.
The company that unblocks a bottleneck may profit, but the giants are built by those who design for the new abundance. And in a world of abundance, profits accrue to whoever can locate and defend the next true scarcity.
Just like the general purpose technologies that have preceded it, AI will need its own complements to reach the mainstream. New form factors beyond chat boxes. New organizational designs that integrate machine intelligence with human judgment. New UX primitives for delegation and supervision. These are bottlenecks too, and until they’re built, AI’s full potential will remain bottled up.
This sort of analysis is particularly urgent because AI is currently removing bottlenecks all over the place. As a result, the most obvious, consensus ideas are getting funded fast and furious. But the uncomfortable implication of the bottleneck cascade framework is that this is a uniquely dangerous moment for incremental bets. They feel safe, but they’re almost always designed for a world that is already slipping away. And worse, they’re often aimed at categories where AI is collapsing margins, leaving little profit even if adoption takes off.
History is full of cascades, each showing how today’s bottlenecks might unfold. The question isn’t whether AI will face them, but what kind of strange, AI-native firms will emerge once those constraints give way.
FROM THE BLOG
No More Painting by Numbers
It’s the end of the “SaaS playbook.
The Age of Artisanal Software May Finally be Over
Every wave of technological innovation has been catalyzed by the cost of something expensive trending to zero. Now that’s happening to software.
Founders as Experiment Designers
David on why founders should run everything as an experiment.
When Growth Stalls
Or why to kickstart growth you should narrow your ICP.
WORTH READING
ENTERPRISE/TECH NEWS
New attack vector. Anthropic published a report that they are seeing cyber attacks conducted entirely by AI agents, with no humans involved. “Agentic AI systems are being weaponized: AI models are themselves being used to perform sophisticated cyberattacks – not just advising on how to carry them out.”
The allure of AI. New research suggests that the people most drawn to AI tend to be those who understand the technology the least. “In one experiment, the researchers recruited 234 undergraduates, assessed their AI literacy and then asked them to consider writing four papers on topics ranging from how the assassination of Archduke Franz Ferdinand led to World War I to a poem about falling in love in Venice. The participants were then asked if they would or wouldn’t use a free version of an AI system to help them complete the assignment and to what extent. The students who scored lower on AI literacy were more likely to use AI to complete the assigned tasks than the students with higher AI literacy, the study found.”
The end of pricing power? Yoni Rechtman of Slow Ventures offers a brilliant analysis of the impact of AI on software margins, the ultimate driver of the attractiveness of a business. “This isn't just about AI companies having negative gross margins (which is obviously bad business model chicanery and likely fleeting). It's about the fundamental economics of software businesses getting permanently worse and thinking through the logical responses. Some of the AI companies currently at low or negative margins will win with capital long enough to survive and raise prices. Some of them will ultimately differentiate through some kind of market depth, nfx, or hybrid models. Many of them will die in the crib.”
HOW TO STARTUP
Unprecedented speed. Bessemer published an update on their Cloud 100, highlighting the speed with which some AI-powered companies are scaling. “The rise of AI-native companies has also fundamentally changed the trajectory to Centaur status, a term we debuted in 2022 for businesses reaching $100 million ARR. In this year’s cohort, it took the average Cloud 100 company 7.5 years to reach Centaur status, the fastest timeline in the list’s history. AI companies are accelerating at an even faster pace, averaging 5.7 years to reach Centaur status (a whole year shorter than AI companies in the 2024 cohort).”
Running towards fear. Lenny Rachitsky published a 90-minute interview with Ben Horowitz of A16Z. It’s full of the sort of hard-won wisdom that made Horowitz famous previously (The Hard Thing About Hard Things), and it’s worth a listen for any founder struggling with the challenges of the moment.
Is your VC giving up on you? Does it matter? Jason Lemkin explained why so many VCs are rotating their attention away from their “pre-AI” slow-growers and towards their newer crop of high-potential AI-native companies. This, he argues, may be just the catalyst some companies need to get serious about profitability and sustainable growth. “VC abandonment forces brutal prioritization. When you know there’s no safety net, you stop playing it safe. You can’t afford to run 47 experiments simultaneously or chase every shiny growth hack. You focus on what actually moves the needle.” The piece is worth reading, containing a lot of practical advice for how to rethink a business in this situation.
Resilience. The Wall Street Journal on Israel’s surprising economic resilience.
HOW TO VENTURE
Consensus capital. Leslie Feinzaig of Graham & Walker VC put out a brilliant piece drawing a distinction between venture capital and “consensus capital.” It is an extremely clear-headed analysis of the state of the venture market, and the best response to Martin Casado’s now-infamous tweet on the dangers of non-consensus investing. There is a role for consensus capital, but there is also a need for venture capital in the classical sense. “Since its earliest iterations in the 1940s, Venture Capital has always meant investing in early stage companies with the potential to generate alpha - high-risk, high-reward, uncorrelated with efficient (public) markets. It’s never been about investing in the obvious. Quite the opposite, in fact. That’s not how big funds invest anymore - Andreessen partner Martin Casado’s viral tweet last week acknowledged this: large funds are not picking contrarian bets. They’re picking consensus ones. All of them are chasing the same founders, outbidding each other in giant rounds, competing away alpha for themselves and each other.”
The Venture Mind Virus In a similar vein, Nic Poulos penned a great piece on the current state of venture capital, highlighting what he calls the “consensus crisis” in venture capital. “The early-stage financing landscape today is flowing into the same handful of themes, founders, and even funds. Valuations are converging upward as the narrative unfolds of winner-take-all markets, where scale is all that matters. More than ever, investors are outsourcing conviction to a handful of signals: Y-Combinator, elite pedigrees, mega-fund co-signs. More than natural acquirers, incumbent technology platforms have their hands on the rudder—with Microsoft, Amazon, and Meta making unprecedented direct investments and talent bids. The warning signs are flashing bright red that the venture market has never been more consensus-driven. We believe that the consequences of continued concentration will be catastrophic for venture capital and the broader innovation economy.”
PORTFOLIO NEWS
Reco achieves industry-leading milestone with 200+ SaaS app integrations.
Beebop is one of The Generalist’s Future 50, a list of the world’s 50 most promising companies.
PORTFOLIO JOBS
Reply