The 2028 Panic and Post-AI governments
Well, rarely has a small firm’s research post gathered so many reads, shares, and sells. What would usually just be a small twitter / substack event, clearly transcended boundaries until even the WSJ covered it, linking it to the drop in stocks. Do public investors really not do any thinking whatsoever for themselves?
Anyways, the post, as I’m sure you’ve all already seen, is here: THE 2028 GLOBAL INTELLIGENCE CRISIS. It is obviously worth a read. (This comes after the “Something Big is Happening” post that was also widely shared, and comes to similar conclusions albeit from a product / coding perspective)
The tl;dr is as follows:
SaaS to zero
Credit cards to zero
Doordash / delivery to zero
Private credit to zero
Mortgages to zero
Real estate to zero
It is in essence quite gloomy.
We’ve long written about our thinking of a post-AI world (and post-robotics), and explored these questions at our Unruly Futures event.
There are quite a few things we disagree on with the piece, especially some moats and which businesses go to zero in such a scenario, but it is undeniable that the amount of thinking on second and third order effects of replacing human labor with artificial intelligences hasn’t been nearly enough in recent months.
The pushbacks can be well summarized by this short article, although some went through the effort of rewriting the full memo in the positive (funnily enough, probably AI written). Manidis tried his own critique, which feel a bit overly intellectualized for most of us, but you can have a stab at it.
The most interesting aspect is understanding if AI brings us an infinite technology-driven deflation, because then the economics change substantially.
Post AI economics should be a constant thought for leaders everywhere. We can debate the timing, but it’s going to be hard to defend the position that “we’ll just figure it out”. And you really don’t want to be one of the simplistic UBI callers (who gets the income, from where, who will still have to pick the tomatoes up, etc. are all very simple questions that no one really has backed-up answers to).
In all of this, we’d really like to point out a few of our favorite quotes in the article
“The federal government’s revenue base is essentially a tax on human time. People work, firms pay them, the government takes a cut. Individual income and payroll taxes are the spine of receipts in normal years.
The government needs to transfer more money to households at precisely the moment it is collecting less money from them in taxes.”
This is really to us the big unexplored area: what are the N-th order effects on governments, and what are the possible actions we can envision right now.
It will merit a post of its own, but it’s not hard to see that governments will have a particularly rough time if unemployment spikes, and tax receipts collapse at the same time (and this will be even more evident for European governments who won’t even be able to tax the AI revenues).
It isn’t particularly outlandish to imagine extreme measures that governments do have in their arsenals: inference bans, data center bans, energy surcharges for computing, nationalization of AI labs, ad-hoc taxes, and so on.
We are already debating some times if we should prioritize machines or humans (ram, computing, energy, real estate, water - are all resources that will be heavily competed)
I think we really need to start projecting a world in which these things are not only possible, but will be cheered on by the vast majority of the population who might no only find themselves without a job, but completely lost as to what do to with the lives.
We will publish a specific piece on our thesis for positioning here soon.
In the meantime, here’s some more awesome stuff we read since the last edition:
AI
Feltsense raises $5M to build AI agent founders, who will then compete with YC to hire the best human operators
Sam Lessin argues we’re watching the biggest capital bonfire in recent memory
What we’ve been saying for years: “The most important businesses will be regulated, capex intensive, operating in the physical world, or have substantial network effects”
Seedance 2.0 has been blowing up, we’ll share one tweet but it’s full of insane examples/
Chinese billionaire founder of trip.com seems to share our worry on depopulation
ElevenLabs v3 says they pass the Turing test for AI voice agents
Simile: human behaviour simulation
Other:
Beautiful, must-read post from Michael at Compound on how now vc-funded startups are low status. Many thoughts here we can’t share yet, but must read.
Works-in-progress figures out inflation is bad for everyone, absolutely incredible to see smart people just figuring this stuff out. Honestly shocking.
Sugar prices in total collapse (as beef skyrockets)
Chinese humanoid soldiers - AI generated or not, this world is coming
Fully autonomous F1-like student cars are 100x better than even 2 years ago