Early last year venture capitalist Aubrie Pagano set out to write a “short white paper” that dug beneath the AI hype in search of a framework for understanding its impact on society and investment.
It turned into something else.
As she and co-author Josh Jagota wrestled with the problem of understanding something so transformative, they wondered if what they were writing was profound—or profoundly unhinged. In my view they’ve achieved the rare feat of compiling an analysis that is both intellectually stretching and practical.
Pagano and Jagota both work for Alpaca VC, an early stage venture capital firm1. Their perspective feels relatable because they aren’t tech insiders, they don’t build AI models, but they do need to understand AI’s potential to thrive in their careers. We might not all be investing in AI, but we’re all invested in understanding how to live in the world it’s creating.
Their paper has two parts. The first analyzes the near-term impact of Artificial General Intelligence (AGI). The second half is speculative, part science fiction and part philosophy. They sketch out the contours of a future society - The Aquarius Economy - and then works backwards, deducing what that might mean for our lives and investment.
Why Aquarius? Astrologically we are leaving the age of Pisces and entering Aquarius, where science, technology and humanist truth-seeking is supposed to dominate over dogma. It will be a challenging age. Human spirit and agency will face the threats of technological overreach and autocracy. But, hey, the authors are VC’s so they also imagine solutions to these tensions and sketch out business “we wish someone would start - to illustrate the kinds of ways we believe the human Spirit can flourish economically even if their traditional comparative advantage is eroded.”
The Last Economic Cycle Built On Labor Scarcity
Pagano thinks AGI will make some cognitive jobs so cheap to do that labor itself becomes infinitely accessible and no longer scarce. This rhymes with past technological revolutions, but isn’t the same thing. Previously, when technology displaced labor it raised productivity, driving faster economic growth. This improved real wages for those still employed and they spent their money on new things - services or products that didn’t exist before. Employment changed, but ultimately grew. I asked Pagano if she thought the same dynamic would play out this time:
We think that’s true in parts, but we also think that creates a lot of peril in the short- and medium-term because it’s not actually clear where, you know, the tens of millions of knowledge workers go.
I don’t know either, but then again I suspect neither did those observing previous technological revolutions. But Pagano isn’t convinced we are near this point. There are a long list of “blockers” impeding the transition to full AGI. Companies that can solve those problems are where the next round of AI fortunes will be made.
Energy abundance is the first category of blockers and the anecdotes here are stunning. AI-related energy demand is so extreme that by 2030 we will need over 150 gigawatts of new power, which is like double California’s entire grid. Anthropic alone will need power equal to four times New York City peak demand. The energy constraint might already be biting. George Robertson of The Monetary Frontier Substack estimates that if current claimed data center build outs were accurate, it would have already generate 500,000 new construction jobs, whereas only about 180,000 have been reported.
Unblocking this will involve some brute force - more power from more sources, including nuclear (and while Trump remains in power, probably coal). All this new power will be a huge strain on the grid, so beyond the opportunities in providing the power and connecting it to the grid, we will need new solutions to improve grid resiliency and optimize power usage.
Robots Can Sing Happy Birthday, But Not Fix Your Plumbing
I was eating dinner at an airport hotel a few months ago when a robot appeared from the kitchen and sang happy birthday to a woman at the next table. Everyone loved it. I assumed the waiters especially so, since it relieved them of this embarrassing task. But getting from happy birthday to replacing labor at scale is a big leap. Here is where the second blockers category emerges. Can foundational industries - manufacturing, supply chain, agriculture, real estate, construction - execute on AGI automation potential in the real world?
AI-powered robots should be good at many tasks in these areas, but data to train them is often unusable or nonexistent. Without good data we can’t build robots to automate the work in these areas. In Pagano’s words:
I just think these have to be solved before you get to these really sophisticated visions that everyone has of these general purpose robots that, you know, fix your plumbing.
With large-scale automation still distant, then the massive shortage of blue collar labor in these industries becomes a blocker too. In my view the obvious solution - the one that has been so beneficial to the US throughout history - is importing labor. But with Trump’s violent and unaccountable immigration police roaming the country, that’s impossible in the short-term.
Pagano’s focus is on the long-term and she sees opportunities in “skilled trade enablement”. This spans education, but also corporate training, including upskilling and reskilling. Education is particularly fascinating as our system from an early age is geared toward prepping for university and white-collar jobs, the very positions that are most at risk. Can there really be a quick pivot to channel kids into the trades? Would white-collar parents even support that?
Can We Talk?
The third blocker category, “agent and human coordination”, might be the most challenging, because leading AI companies probably don’t want the issues unblocked. Coordination refers to the effectiveness of human-to-agent and agent-to-agent communication and cuts to the heart of big tech’s business model: they want to create walled gardens. Their initial product is good, both in terms of quality and cost, but once you’re in the garden and the gate is locked, queue up Enshittification - the erosion of product quality that we all now suffer. Pagano knows the problem well:
If you want to go between ChatGPT and Claude and have them interact, they’re totally walled gardens, the memory layer between them gets lost. If I want to transfer something over, I can’t.
And so, as you’re trying to build, with these tools, more sophisticated workflows and actually trying to automate away real tasks, you run into this clunkiness that we call coordination, that ends up blocking.
Putting all the blockers together, Pagano thinks we may be two or three capital cycles away from full AGI.
The Aquarius Economy
In the paper’s second part, she assumes all the blockers are solved and tries to depict what society looks like. This is the science fictiony bit, tinged with some philosophy. Still, she keeps her VC hat on, with an eye to the roles, services and products that will be most highly valued in this world. She admits this is speculative, but her process felt intuitive to me: start with a generalized view of what the world might look like in 20, 30, 40 years time and work backwards to see what type of long-term investments that might create.
Power in this world rests with two “groups” - the hegemony and the technocore. The hegemony is the small number of uber wealthy families and corporations that control the economy’s resources. The “technocore” are the centralized AGI systems that run (most) of the world.
Not everyone will buy into this future, so she envisions Nomads who opt out of centralized systems and live, as much as they can, off the grid. They are tolerated by the centralizing powers as long as they don’t challenge them. “No harm, no foul” as Pagano says.
Gurus representing the apex of human spirit - artists, athletes, healers - will occupy an exalted space in this tech-centric world, because their lives most embody the one thing AGI can’t emulate - the human spirit. I chuckled when, a few days after interviewing Pagano, The Athletic website sent me an overview of their new Peak Newsletter where readers get “actionable insights and lessons on performance, mindset, and routine” from elite athletes.2
And lastly there are the Incels, socially and spiritually disconnected individuals. We’ve already seen this group emerge, but they will get larger. From an investment perspective, Pagano’s framework is a kind of lens she uses to identify early investment opportunities in human connection, trust, care and reputation.
It’s a disservice for me to summarize the entire post-AGI world in three paragraphs! For a full picture download the whitepaper from Aubrie’s Substack.
Listen to us discuss her ideas on Top Traders Unplugged Ideas Lab podcast series:
Spotify:
Apple Podcasts:
Full disclosure: I’m an investor in several Alpaca VC funds but I’d never met Pagano before our podcast interview. Nothing in this post, or in our conversation, refers to specific Alpaca VC investments or managed funds.
such a great interview and white paper - READ it