You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
"I was the bottleneck, not the AI" — David Shapiro
An unstructured ramble covering OpenAI's O3, the nature of AGI, cognitive horizons, and a first preview of post-labor economic theory. Recorded April 18, 2025.
This is a valid approach — many breakthroughs are novel combinations of established ideas rather than entirely new inventions.
Comparison to Existing Proposals
vs. Universal Basic Income (UBI)
Aspect
UBI
Shapiro's Post-Labor Economics
Income source
Government transfers
Property/ownership dividends
Funding
Taxes, money printing
Local resource monetization
Agency
Passive recipient
Active co-owner
Scalability
National policy required
County-level, bottom-up
Dependency
Increases transfer dependency
Shifts away from transfers
Political feasibility
Highly contentious
Builds on existing models
Shapiro explicitly positions his model as anti-UBI — UBI increases the "transfers" component of EAI, while his model aims to increase the "property" component. However, UBI could serve as a bridge during transition.
vs. Georgism (Land Value Tax)
Henry George's proposal to tax land values and distribute proceeds has significant overlap:
Both recognize land/resources as communal wealth
Both aim to distribute returns from natural resources
Shapiro's model is more granular (county-level) and doesn't require new tax legislation
vs. Stakeholder Capitalism
Similar to stakeholder capitalism models (B-Corps, ESG investing) but more radical — actual ownership stakes for residents, not just "consideration" of stakeholder interests.
vs. Community Land Trusts
Community Land Trusts (CLTs) already exist and do something similar at a smaller scale. Shapiro's model is essentially CLTs + co-ops + sovereign wealth funds, systematized with metrics.
Potential Counterarguments & Weaknesses
1. The "Who Builds It?" Problem
Co-ops and infrastructure require initial capital. Poor counties with low CPP can't bootstrap themselves — they need external investment. Shapiro doesn't address how the first dollar gets invested in a county that has nothing. The Alaska Permanent Fund was seeded by oil revenue — not every county has an equivalent windfall.
2. Governance Capture
County-level governance is notoriously susceptible to capture by local elites. Pushing economic decisions to counties could empower local oligarchs, not residents. Small-town politics are often more corrupt than federal politics, not less.
3. The Mobility Assumption
The model assumes people can freely relocate to optimize their dividends. In reality:
Housing costs create lock-in
Family ties, schools, healthcare access constrain mobility
Moving costs are disproportionately burdensome for low-income individuals
Immigration status restricts movement for non-citizens
4. Race to the Bottom
Counties competing for residents via dividends could create perverse incentives:
Overexploiting natural resources for short-term payouts
Lowering environmental standards to attract industry
Excluding "undesirable" populations to maximize per-capita returns
5. Property Concentration Paradox
Shifting from wages to property income assumes property can be broadly distributed. But property already concentrates toward the wealthy. Without strong redistribution mechanisms, post-labor economics could accelerate inequality rather than solve it. The top 10% own ~70% of all wealth — co-ops alone may not counter this structural concentration.
6. Scale & Complexity
There are ~3,143 counties in the US. Creating individualized economic strategies for each, with dashboards, metrics, and intervention menus, is an enormous governance and data infrastructure challenge. Who maintains this? Who enforces accountability?
7. The "Third Metric"
Shapiro mentions a third unnamed metric. Until revealed, the model is incomplete. This could be the keystone or a hand-wave.
What's Genuinely Interesting
Despite the above critiques, several ideas have real merit:
Measuring what matters — The insight that wages/property/transfers ratios aren't currently tracked at a granular level is correct and important. "What gets measured gets managed" is a genuine lever.
County as economic unit — Most economic policy happens at federal or state level. County-level economic strategy is underexplored and potentially powerful — counties are small enough for local knowledge, large enough for meaningful economic activity.
AI as economic research accelerant — The meta-narrative (using O3 to develop economic theory in an afternoon) is itself a powerful demonstration. Whether the theory is right or wrong, the process of rapid theory development with AI is genuinely new.
Self-correcting population distribution — The carrying-capacity model for counties, where overcrowding naturally pushes people to underpopulated areas, is an elegant market-based mechanism. It's reminiscent of "voting with your feet" (Tiebout model, 1956) but with a concrete dividend mechanism.
Sustainability incentive alignment — Tying local ownership to local resources genuinely does align incentives for sustainability. You don't overfish your own fishery. This is well-supported by Elinor Ostrom's work on commons governance (Nobel Prize, 2009).
Open Questions
How does this interact with existing property ownership? Homeowners already have property-based economic agency. Does the model double-count?
What happens to counties with no clear "unfair advantage" — flat, dry, no resources, no scenery?
How do you prevent wealthy outsiders from buying up co-op stakes and recreating the current wealth concentration at county level?
Does the model account for automation of the management of these resources? If AI runs the solar farm and the fishery, who's left to "work" at the county level?
How does international competition affect county-level economics? A county's solar co-op competes with Chinese solar manufacturing at global scale.
Cognitive Horizons — The totality of what a mind is able to mentally represent. Expands with knowledge, experience, and education. AI acts as an accelerant to expand these horizons faster than traditional learning. (Shapiro's prior videos)
Cognitive Offload — Delegating thinking to AI without internalizing the insights. Shapiro warns this leads to intellectual atrophy — "BSing your way through meetings with AI."
Intelligence Blast Radius — The scope of problems a model can usefully address. O3's blast radius is dramatically wider than prior models.
Plastic Intelligence — AGI's ability to become good at anything without the trade-offs organic brains face (e.g., dyscalculia, dysgraphia). The entire cognitive stack is reconfigurable.
Frontier of Automation — Term from Anton Korinek at the IMF, describing how technology expands the sphere of what humans are cognitively capable of. (IMF working papers)
Multi-Intelligence Theory — Intelligence as a multi-dimensional landscape (not just IQ). Hundreds of dimensions; human brains are "lopsided" due to genetics, while AGI can optimize any dimension.
3.0 — Silicon substrate, entirely plastic stack (AGI in any form factor)
S-Curve of S-Curves — AI progress isn't one sigmoid but stacked sigmoids — each paradigm (scaling, synthetic data, distillation, reasoning) creates a new growth curve before the previous one saturates, producing what looks like a step function.
Distillation — Compressing a larger, smarter model's capabilities into a smaller, faster, cheaper model. Shapiro believes O3 is a distilled version of O1.
Post-Labor Economics
Economic Agency Index (EAI) — A composite metric measuring an individual's (or county's) income composition across three categories:
Wages — Money from exchanging labor
Property — Passive/residual income (stocks, bonds, rental, co-op dividends, IP royalties)
Transfers — Government distributions (Social Security, UBI, food stamps, Medicaid)
Goal: shift the ratio from wages → property over time.
Collective Purchasing Power (CPP) — How much capital a county can mobilize without increasing taxes or debt. Higher CPP = more reinvestment capacity. Highly correlated with EAI.
Aggregate Demand — The total amount people are willing and able to spend. If wages collapse due to AI automation, aggregate demand collapses → economic crisis. This is the core problem post-labor economics tries to solve.
Subsidiarity — Push every decision (governance or economic) to the lowest level of governance capable of making it. In post-labor economics: county-level interventions, not federal mandates. (Wikipedia: Subsidiarity)
Measurements and Menus — Economics as diagnosis + prescription. Measurements (GDP, EAI, CPP) provide diagnosis; menus of interventions provide prescriptions (co-ops, sovereign wealth funds, infrastructure investment).
Financial Outpatient — Term from The Millionaire Next Door by Thomas Stanley & William Danko. Someone whose livelihood is heavily subsidized/propped up by external support (government transfers). Counties with high transfer dependency are "financial outpatients."
Unfair Advantage — Every county has unique assets (rivers, solar potential, fisheries, farmland). Post-labor economics leverages these via co-ops and local ownership.
Co-op (Cooperative) — Collectively owned infrastructure/utilities. Example: a solar co-op in Arizona where residents receive monthly dividend checks from energy production.
Carrying Capacity — Natural population equilibrium for a county. When too crowded, per-capita dividends decrease → people migrate to underpopulated counties with higher returns. Self-correcting distribution.
Dividend Tourists — People who move to a county solely to collect dividends then leave. Mitigated by vesting schedules (the longer you stay, the more you're invested).
Brain Drain — Loss of talent from a region. Post-labor economics disincentivizes this by tying economic interests to local land and resources.
Circular Economies — Economic systems where resources are reused/regenerated rather than extracted. Locally-owned resources incentivize sustainability because over-extraction stops dividend production.
Alaska Permanent Fund — Real-world example of a sovereign wealth fund that distributes oil revenue dividends to Alaska residents. (Wikipedia)
Cross-References & Further Reading
People Mentioned
David Shapiro — AI researcher, content creator. YouTube | Linktree
Max Tegmark — MIT physicist, author of Life 3.0. Wikipedia
Anton Korinek — Economist at the IMF, work on AI and the frontier of automation. Website
Thomas Stanley & William Danko — Authors of The Millionaire Next Door
pie title Income Composition (Current vs Goal)
"Wages (labor)" : 60
"Property (ownership)" : 25
"Transfers (government)" : 15
Loading
Goal: Shift ratio over time so Property dominates, Wages shrink, Transfers decrease.
Post-Labor Economics Framework
flowchart TD
subgraph Problem["Core Problem"]
AI[AI & Robotics Automation] --> WC[Wage Collapse]
WC --> ADC[Aggregate Demand Crisis]
ADC --> EC[Economic Collapse Risk]
end
subgraph Measurement["Measurements (Diagnosis)"]
EAI["Economic Agency Index (EAI)<br/>Wages | Property | Transfers"]
CPP["Collective Purchasing Power (CPP)<br/>Capital mobilization capacity"]
M3["Third Metric (TBD)"]
end
subgraph Principle["Guiding Principle"]
SUB["Subsidiarity<br/>Push decisions to county level"]
end
subgraph Interventions["Menu of Interventions (Prescriptions)"]
COOP["Co-ops<br/>(solar, infrastructure, utilities)"]
SWF["Sovereign Wealth Funds<br/>(Alaska Permanent Fund model)"]
ESOP["Employee Stock Options"]
INF["Infrastructure Investment<br/>(fiber, power, greenways)"]
end
subgraph Outcomes["Desired Outcomes"]
PROP[Property income grows]
SUST[Environmental sustainability]
POP[Population redistribution]
BD[Brain drain reduced]
end
Problem --> Measurement
Measurement --> Principle
Principle --> Interventions
Interventions --> Outcomes
Outcomes -->|Feedback: higher EAI & CPP| Measurement
style Problem fill:#ff6b6b,color:#fff
style Measurement fill:#4ecdc4,color:#fff
style Principle fill:#45b7d1,color:#fff
style Interventions fill:#96ceb4,color:#fff
style Outcomes fill:#a8e6cf,color:#000
Loading
County-Level Feedback Loop
flowchart LR
A["Identify County<br/>Unfair Advantage"] --> B["Build Co-op<br/>(solar, fisheries, etc.)"]
B --> C["Residents Get<br/>Dividend Checks"]
C --> D["EAI Shifts:<br/>Wages → Property"]
D --> E["CPP Increases"]
E --> F["Reinvest in<br/>Infrastructure"]
F --> G["County Becomes<br/>More Attractive"]
G --> H["Population Grows<br/>(to carrying capacity)"]
H --> A
style A fill:#f9ca24,color:#000
style B fill:#f0932b,color:#fff
style C fill:#6ab04c,color:#fff
style D fill:#22a6b3,color:#fff
style E fill:#30336b,color:#fff
style F fill:#535c68,color:#fff
style G fill:#6ab04c,color:#fff
style H fill:#f9ca24,color:#000
Loading
Population Self-Correction Mechanism
flowchart TD
OC["Overcrowded County<br/>(e.g., LA County)"] --> LD["Per-capita dividends<br/>decrease to $500/mo"]
LD --> MIG["People migrate out"]
UC["Underpopulated County<br/>(rural, resource-rich)"] --> HD["Per-capita dividends<br/>high at $2500/mo"]
HD --> ATT["Attracts new residents"]
MIG --> UC
ATT --> |"Over time"| OC
EQ["Natural Equilibrium:<br/>Carrying Capacity"]
UC --> EQ
OC --> EQ
style OC fill:#e74c3c,color:#fff
style UC fill:#2ecc71,color:#fff
style EQ fill:#3498db,color:#fff
Loading
Cognitive Horizons & AI as Accelerant
mindmap
root((Cognitive<br/>Horizons))
Without AI
Years of study
Limited by biology
Trade-offs (dyscalculia, etc.)
Single-domain depth
With AI (O3)
Afternoon learning
Stress-test ideas in real-time
Cross-domain synthesis
"Brain is the bottleneck"
Risk
Cognitive offload without internalization
Intellectual atrophy
"BSing meetings with AI"
AGI Trajectory
Life 1.0: Biological only
Life 2.0: Biological + Neural
Life 3.0: Silicon, fully plastic
i just woke up and it's p uh pollen season so if my voice is a little bit scratchy just please forgive me and bear with me um but uh I got some feedback that some of you guys really liked my more kind of unstructured rambles that you got a lot of insight out of it so let me know what you think i mean the numbers will tell me one thing but you know let me know in the comments anyways I wanted to talk about OpenAI's 03 full uh the nature of AGI and post labor economics um that's kind of the extent of the structure that I've got so let me just press play and start rambling
so first and foremost 03 came out a couple days ago and I have been using it uh I'm not going to say non-stop but for several hours every day um and there is something very qualitatively different about this model uh if you just look at the benchmarks it looks like it's incrementalism if you look at uh the hype cycle it looks like typical hype typical hype there we go not typical hype it looks like typical hype um but at the same time I think it has crossed a threshold so many many technologies in the past have crossed uh thresholds at at at certain times
um you get you you basically get to uh one one of the key things in economics is you get to network effects um but there's also just various tipping points i kind of think of it as like a utility function tipping point where okay this technology goes from you know novelty and cute and quaint to actually this is this is really useful um and it's been a it's been a sliding scale like obviously chat GPT has been useful for some things uh all along um but its intelligence blast radius was relatively limited um you know it would still hallucinate there's stuff that it didn't know it would you know lie to you all all kinds of problems
i think that 03 is the model that has that has really gone past the goalposts u in terms of like yes this is general purpose utility for pretty much everyone um now is it perfect at everything no does it is is it going to win the the gold medal at the you know international math olympiad the IMO no not yet but in a generation or two it will and anyways you don't need to win the gold medal at the at the math olympiad to do groundbreaking research um that's if you want to be the best mathematician in the world sure but it's already at that level with coding and and general knowledge like the the the was it the Google proof the GPQA diamond right it's already at you know idetic memory right and knows everything and doesn't hallucinate too much to be useful
so um I've been going bonkers on on Twitter about 03 and what I'm using it for um so the three primary categories that I've used it for just I'm not going to like bore you to death with all the threads but is philosophy literature and post-labor economics um and then also for some of my own health and healing uh which by the way uh it is it is medically the most superior model that I have used this is not medical advice it's like I'm not saying like you know fire your doctor and use 03 but I mean I kind of have so don't do do as I say not as I do
um but yeah so I updated uh all of my you know medical information in my my chronic fatigue burnout uh project and 03 uh you know with the new memory feature as well it has a broader picture of your health so I'm like "Okay let's just take a step back new model." I said "Look look at my look at my health and just let me know and and everything that we've talked about let me know what you agree with let me know what you disagree with like what have we gotten wrong up to this point?" And it went through and it said "Okay all this is correct here's a couple things that you know need a little bit of correction or we're not quite right." Nothing was catastrophically wrong so the previous generation of models uh from you know 03 mini doing deep research to uh GPT4.5 which I understand why a lot of people loved GPT 4.5 but honestly 03 is smarter and many times faster so I'm not sad that GPT 4.5 is going to go i'm glad I got a chance to use it um fortunately OpenAI upgraded me to the research preview at least for that model i don't have access to all the the research stuff but I did have I did have more access to 4.5
um and so anyways 03 is smarter and faster so you know and and it's also very very flexible um whatever you want 03 to do it can do it can write code it can write pros it can do philosophy it can do economics it can search the internet whatever you want it to do it can do um so yeah burnout it and the good news there is that um I'm actually further along in my healing trajectory than I thought it's that I've got about six to eight more months give or take you know natural natural variance it said plus or minus 30 to 50% um which is better than the 18 months of healing that I thought that I had left anyways all right that's enough of that
so philosophy literature post labor economics 03 can do it all um it's fast it's uh it's actually cheaper as well so I think that 03 is probably the highly distilled version of 01 um so not only is it cheaper and faster it's smarter so the next generation basically what we're going to be waiting from here on for um 04 which maybe 04 is just another distilled version of 03 i don't know i'm not really sure what the training cadence is because the what we what we thought or what I thought was happening is they were going back and forth between using synthetic data to build up a corpus and and do a training run on a on a next next you know scale uh frontier model and then use distillation to compress that ability make it smaller faster cheaper but it seems like 04 is on deck and 04 is even smarter than 03 so I'm not sure what they're doing behind the scenes
um but the the training cadence seems to be picking up it's you know I I remember there was a lot of you out in the comments that said you know like yes every technology is an S-curve but it's like an S-curve of S-curves and I thought you guys are being ridiculous that's not how technology works but I'm I'm actually at least in this particular niche um because if you take a big step back that's all of science is slowing down all of technology is slowing down um but at least with AI we are seeing like exponential and then another exponential or another S-curve so sigmoid sigmoid sigmoid so you have basically it looks like linear step functions right so it's like yes the you know one paradigm might might burn out but then you get another paradigm and then you get another paradigm so it's almost like instead of diminishing returns it's almost like a linear function where there's you know sigmoid exponentials or brief exponentials but really sigmoid uh as a step function anyways okay so that's rambling
AGI: Cognitive Horizons (7:05)
um I want to pivot to AGI so I have long held like AGI September 2024 um it it really is looking like the the reasoning models are the proto AGI now yes I will be the first to see that there's lots and lots of stuff that it can't do yet um you know it's got relatively short time horizons but at the same time like these models are are capable of things that that most of us are not capable of
um and so I wanted to talk about that because I've been I've been thinking honestly like okay 03 is the first model that I have used that it feels like it's doing things that I legitimately cannot do but it's not an alien intelligence because it's like okay you know here's all the here's the constellation of dots it it did it faster than I could you know just as an example I've been working on post-labor economics for a couple years now and in an afternoon it searched the the internet for all of my previous work on post labor economics used its knowledge about everything else and said "Here's the metrics that you need here's how to measure it here's uh here's what what the theory is let's unpack
And I felt like literally when I was every every conversational turn every dialogue turn that I used with 03 while working on post labor economics felt like a day's worth of work with deep research um like like okay so in in like two or three dialogue turns with 03 we figured out okay here's the data sources that we need cuz all the all the data that we need to create an Economic Agency Index which is the beating heart of post labor economics all the data is out there here's the data here's what you need to do to clean it here's how to make the calculation do you want me to write the the Python notebook right now and I'm like hold on it's like I'm like hold on
so my brain is now the bottleneck In the past the intelligence of the AI was the bottleneck no matter how smart set 3.7 was or 01 it's like it would it would be really random like going off in the wrong directions and I'm like "Hold on i got to think about this you you're you're I I was the one keeping it on course but now I've set the direction and and the AI is like "Come on hurry up hurry up you dumb ape." Um so it's it's not talking to me like that but that's that's the sense that I got where it's like my velocity my cognitive and intellectual velocity is now the constraint and I'm pretty smart
so for me 03 has been a tipping point where it's like yes you know chat GPT could write pros faster than I could it was in the grand scheme of things it's pretty generic pros it's good it's it's decent pros it's passible it'll tell a story but it could write the pros faster than I could just because synthesizing pros i'm limited by the speed of my fingers and and and those sorts of things um you know it could code faster than I could everyone has known that but you know it it wasn't solving problems necessarily that you couldn't solve it just put it together a little bit faster but in this case even even drinking in the output of the AI as fast as I could I'm like hang on I still have to internalize this um so over the course of a few hours yesterday we hammered out all of post-labor economics and I'll get back to that in a in a second
um but so really what it seems like AGI is doing is it's not I so all right so having studied intelligence for for a long time now um because I wanted to understand intelligence and model it with cognitive architectures and those sorts of things there is there there's cognitive horizons which I've talked about in previous videos and the short version is your cognitive horizon is this is the totality of what you are able to mentally represent
um now your cognitive horizon can expand so like for instance even if you're a clever 10-year-old you don't necessarily know enough about history and physics and and calculus and those sorts of things there are things that you cannot mentally represent when you're 10 years old and there's things that you that that you can by the time you're 20 but even still there's things that that by the time that you're 20 you can't represent because you haven't had enough life experience and more education but then by the time you're 30 or 40 your cognitive horizons have not quite peaked but they're so big that there's so many things that your brain has learned to generalize and and to represent
and so the the question about AGI as to whether or not it has an alien intelligence is is the cognitive horizon is the is the functional effective cognitive horizon of a machine greater than a human or greater than all humans right now I still haven't seen anything that that tells me that it's a truly alien intelligence um the reason is reality we we are all operating in the same sandbox math is the same physics is the same now obviously the the substrate the the the cognitive substrate of the of the AI is different but who cares right it's writing the same code it's using the same language it's using the same math as us
so as long as as long as as long as that remains true then you know it's like the frontier of automation which is the the term used by Anton Corn over at the IMF like he pointed out pretty pretty simply technology has usually expanded human cognitive horizon horizons he used a different term but we're talking about the same thing basically the the the sphere of what humans are cognitively capable of typically is expanded by technology
now for the longest time I have believed and I know that a lot of you have believed as well that maybe we were already at the limits of human cognitive horizons and that AI was just going to go sailing past that however after using 03 for for a couple days now I'm not sure that I believe that anymore um I it has certainly expanded my cognitive horizons and it can teach you very quickly um and and one of the one of the trade-offs though is that like rather than rather than spending you know years at college learning you know learning all of these economic principles by wrote you spend an afternoon getting a good enough understanding and asking the model to stress test your understanding as you go so that you're internalizing those ideas uh whether it's medical ideas or economic ideas or mathematical principles or computer science principles or physics principles whatever you can stress test your ideas as you go and that expands your cognitive horizons
so is the cognitive horizon of 03 beyond mine absolutely right now it can do math and coding and and all kinds of stuff that I cannot do right now but what I'm finding is the speed at which I can learn you can accelerate those cognitive horizons pretty quickly so so it remains to be seen as to whether or not AGI will be superset and and and and supersede all human cognitive horizons or if it'll actually just help us all expand our cognitive horizons
obviously every every brain is different there are certain things that my brain is just not as good at timing and rhythm is just I I just don't have timing circuits in my brain so no amount of AI is going to fix that and for some people like you know if you have disgraphia or disalculia like you're just never going to be lexically intelligent or you're just never going to be uh mathematically intelligent um and but you know you're going to have other strengths elsewhere um often people with discalculia or disgraphia uh tend to be more expressive um better with like things like acting and singing and those sorts of things our brains often have trade-offs
now what I will say is that AGI probably doesn't have those trade-offs anything that the AGI wants to get good at or needs to get good at it's completely plastic and so what I mean by plastic is it's like okay well you know just using benchmarks as benchmarks as an example it's like okay we're optimizing for math and then it gets really good at math and then it gets really good at coding but then you know open AAI has been neglecting literature and pros so now they're doing a literature and pros model and so then it'll get really good at that and now it's getting and now it's getting good at images and then it'll get good at video and audio and music and those sorts of things
so you know you can you can you can you can visualize intelligence like the multi-intelligence theory is or is it's like a multi-peak theory or it's a landscape so it's not just like one dimension it's not just IQ there's dozens i mean some some theories of intelligence have hundreds of dimensions of intelligence but human brains because we have an organic substrate that is uh largely I'm not going to say dictated but but strongly shaped by our genetics means that you're going to have a lopsided intelligence landscape in terms of things that you're just naturally better at saying that things that you're naturally curious about agi will not have those constraints
and this goes back to Max Tegmark's uh book Life 3.0 now I know that you'll probably be saying like Dave you're ultra skeptical of of Max Tegmark because he signed the pause letter and I will say yes um a broken clock can still be right twice a day and Max Techmark did contribute to uh in a very meaningful way uh with his book Life 3.0
so basically life 1.0 is biological substrate and that's it so you're thinking amiebas trees right no brains but then life 2.0 is once life evolved organic computers big brains which were far more flexible far more adaptive than just bio uh than just pure biological substrates so then you have a neurological substrate but life 3.0 is when you have a silicon substrate and the entire stack is completely plastic um so what that means is like you can put AGI in anything you can put it in a humanoid robot you can put it in a car you can put it in a data center and then also it can control its own models its own data you know all kinds of stuff that's that's one one way of thinking about AGI
um so in terms of so going all the way back to that time compression idea what it really seems to be doing right now is because the question of whether or not it supersedes our cognitive horizons it's an accelerant um and so when you think about like okay what what is it that it's actually doing what's the value ad uh it's connecting dots for us faster than than we could like it's a timesaver fundamentally it's a timesaver even to this point I don't know that 03 has done things that I fundamentally and am incapable of but it has certainly saved me literally years of time at this point um so I have learned stuff that would have taken me otherwise years to go glean i have figured out stuff that otherwise would have taken me years to go figure out so functionally it's done stuff that I couldn't do just because the time investment was just not really feasible
um yeah so all all that being said I think that we're starting to get a picture as to what AGI actually will look and feel like um and if you don't so just taking a step back you don't have to use this to make yourself smarter you can just do cognitive offload and say "Fig it out for me i'm not going to internalize any of this information i'm not going to internalize these mental models just figure it out for me." You'll be doing you'll you'll be hurting yourself more than helping yourself because then you're not going to be learning it's kind of like um you know like the stories of people of toxic people at work that just always offload their work and then try and pass other people's work off as their own but then it's like they don't actually understand what they're talking about they're just like "Hey I just you know learned enough just to my way through this meeting." Um pardon my language so if you're just BSing your way through meetings with the help of AI you're hurting yourself more than you're helping yourself use these models to internalize that knowledge to internalize those mental models and you will expand your cognitive horizons
Post-Labor Economics (18:48)
okay finally I'll ramble at you about post labor economics um I can imagine some of you will glaze over so if you end the video at this point that's fine um post labor economics comes down to a few things that we figured out so number one measurements and menus uh uh economics is all about measurements and then based on those measurements that you get so like GDP or yield curves or bond market rates and all this other stuff um so that's the measurement right and there's all kinds of other aggregate measurements there's Jenny coefficients and those sorts of things so you have a measurement and then you have a menu of interventions or options you say okay based on this measurement this diagnosis tells us you know these are the knobs that we can we can change the treasury rates or we can you know change so and so and uh you know for instance if if unemployment is too high or too low you know there's things that you can tweak in the economy to change the to to move that needle so measurements and menus is is kind of the name of the game
so what are the measurements and menus that post labor economics is going to offer and this is a very this is a world first early preview of the mostly complete post labor economic theory i do still have a lot of work to do there's a lot of numbers to crunch um but the data is out there and 03 really really is raring at the bit to go get the data for me and crunch the numbers
um so we'll get to that but the the primary question of post labor economics is there is broadening consensus at everyone from Brooking in Brookings Institute to the IMF to the World Economic Forum they're all talk and then you know the UK EU US they're all talking about it seems like labor is going to become a smaller component of how money and wealth is distributed in the future tlddr wages from labor are going to collapse it's the everyone's reading the writing on the wall even the big dogs at the top of the at the food chain they're saying this is going to break down
so how do we shore up aggregate demand so aggregate demand is basically uh is is people's ability or you know the the the the total amount of money that is being that people are willing to put into the market to purchase goods and services because the entire economy right now the the the whole economic theory is that we have a consumption-based economy which is basically uh you need a market and and everything is driven by supply and demand well if you don't have wages you don't have the ability to demand anything because demand is not just what you want and need demand is your ability to pay for what you want and need so if you lose the ability to pay for what you want and need then well then the market collapses and the whole economy collapses
so it's good this is some of the research that 03 was helping me do it's good to see that literally every big dog institute out there from Goldman Sachs on down is is worried about this and none of them have a complete solution they're talking about it they believe the problem they don't have a solution
okay so this goes back to what I alluded to earlier Economic Agency Index economic agency is is an individual metric so it's it's looking at each individual and saying what level of economic agency do you have and economic agency is it's broadly your ability to influence your your economic destiny uh workers rights property rights those sorts of things however what we if you're already working in a in a free market economy in a in a in a liberal society um with capitalist society then a lot of economic agency is already um is already baked in right you know if you assume that you have property rights if you assume that you have social mobility because like I don't I can just up and move to any state that I want to i don't need anyone's permission i can quit a job that I want to anyone can hire me so I already have a lot of economic agency just by by dent of the fact that I'm an American i can buy a house i can sell a house the only per the only permission that I need is from the bank and the bank wants to you know wants to sell me as much as they can because then I'll be paying mortgages or you know auto loans or whatever
um so I already have a lot of economic agency but the real key the the core heart of post labor economics is wages as a component of income and aggregate demand is is going to be going down okay great so how do we shore that up and the the the obvious thing the obvious thing is UBI universal basic income um now I've been talking about an investment-based future or a property based future and I've been experimenting with things like blockchain and those sorts of things
so the long story short is that there are three main categories that are already tracked so I was kind of reverse engineering a lot of work that a lot of that previous economists economists have already been doing and there's a lot of information already being tracked so in terms of economic agency which is basically how much money do you have to spend um because that's that's the neoliberal bargain is your your your economic agency is directly tied to your wallet as long as you've got money you can do whatever you want go forth and be free free market baby right you know greed is good
um so there are three primary components to economic agency that is wages which is money you get from work uh exchanging labor for money there's property which is everything from rental property to stocks and bonds and those sorts of things so that's the that's the passive income or the residual income it can be ownership stakes and businesses it doesn't even necessarily have to be fully passive but it's you the more you own the more you you get back and then there's transfers so transfers is stuff like uh government handouts that's going to be uh entitlements um so you know your social security that's going to be uh UBI food stamps um any any kind of any kind of money that is just distributed to you from the government so those are the three categories wages property and transfers
so the Economic Agency Index of post labor economics is a composite of those three metrics um and so you just look at the you look at the proportions you look at the ratios of those uh numbers and you say okay uh if if most of your income comes from property or passive income or you know residuals from IP like maybe you owned and sold some patents you're already post labor um if you're not if you're not exchanging your time for money then you're already post labor you're fine um if you if most of your income is coming from wages you're not post labor
if if your life if your lifestyle or your livelihood is heavily subsidized by the government already which many of people are through everything from entitlement spending to uh like such as Medicaid and uh social security and food stamps and and rent assistance and all those other kinds of things then you're what uh what they called in um in uh in Millionaire Next Door you're a financial outpatient so you're being you're being propped up by the system and that can come from uh that can come from local government it can come from state government excuse me and it can come from federal government apologies for the pollen we're having one of the worst pollen weeks uh hopefully before it tapers off here in North Carolina
so what you do then is the Economic Agency Index is an aggregate it's a score based on those three metrics wages property and transfers and what you really want to see is you want to see that uh the ratio switched to favor property over time wages are going to go down that's just that's it seems like a foregone conclusion yes there will always be some jobs to go around there's always going to be something for humans to do um AI and robotics might increase uh create some new job sectors like hey I I'm a I'm a creator i'm an influencer on the internet this job hopefully won't go anywhere but if you're a coder if you're a lawyer if you're a doctor a lot of those jobs might go away or be dramatically diminished so we can't really count on those kinds of you know doing doing wages for labor uh to shore up aggregate demand because if everyone loses their jobs the the jig is up and everything crashes
so by by first creating this metric you know because there's that old saying that which gets measured gets managed right now this Economic Agency Index that that ratio of wages property and transfers isn't getting measured so well what you do is you measure it and you measure it at the county level you can you can measure it at multiple levels you can measure it at the county level uh you can measure it at the city level the state level and the national level
but here's the thing the reason that you want to measure it at the county level is the principle of subsidiarity which I've talked about previously in post-labor economics the subsidiarity basically says push every decision whether it's whether it's uh governance or or economic push every decision to the lowest level of of governance that is capable of making that decision now the reason that you want to do this is because then you create a decentralized hive mind you create uh you you you you make use of what the market already allows which is don't make central decisions we don't want socialism we don't want communism we don't want you know Soviet or Chinese style communism and I know some people like "It's not actually communism." Whatever we don't want to operate like the Soviet Union and China okay that's the point i'm not going to not going to argue over semantics we don't want to operate with strong central authority we want subsidiarity we want decisions pushed out to the periphery
because guess what the people who know the most about your county or your city are the people who live in your county or your city and the people who have the most vested interest in making decisions economic or otherwise are the people who live in that vicinity so subsidiarity is a well-established principle um now often subsidiary has to do with with uh elections and voting and and civic decisions but it can also be economic
so we're what I'm going to be proposing and what what the the data that I've got to crunch is I've got to create a a nationwide dashboard of every county because the data is out there um with an economic agency heat map and what that will do is if if I'm right then that will that will provide the measurement which will then provide the diagnosis so the diagnosis is okay okay so we look at county X over here that has low economic agency because you know so low economic agency would be um has high dependency on transfers um and then you know I think that's the primary thing is like high dependency on transfers and then probably um probably higher wages than property
so these are if it's if it's if it's a if it's a county that has high wages that most of the income comes from wages or transfers but doesn't come from property then you know that they're not ready for post labor economics they're already a financial outpatient of the government they're entirely dependent on work if the work goes away that entire county collapses so what you want to do is you want to you want to start tipping that that county away from wages towards property and away from transfers
so that's government subsidies to property now I talked about menus so that's a measurement diagnosis and then there's prescriptions so a prescription or intervention is what can a county do and with the help of 03 there's actually a lot that counties can do and I'm not going to get into the into the litany of details um but it depends on the on the strengths and weaknesses so every county has an unfair advantage
one example could be let's imagine that you live in a county that has a river going through it in most rural counties today rivers are ignored they're just treated like obstacles for highways but what if you say "Hey we can actually make this county more attractive by putting in some green waist trails and other other programs like a a canoe launch." There's all kinds of things that you can do and then so when you put in those things you say this will actually increase the aggregate uh demand for this county this county becomes more valuable well so then what comes with it then you get more infrastructure you get uh fiber optic and internet then you get more power infrastructure and so what you do then is you make sure that each of those um each of those new pieces of infrastructure and public uh public utility uh is collectively owned by you by the by the people who live there as a co-op co-ops are very well established
so then imagine that you're in a county out in Arizona right or you got plenty of sunshine right so you got you know Arizona New Mexico uh they have the most uh cloud-free days so their unfair advantage is solar so instead of instead of building a a greenways trail beside a river that doesn't exist out in Arizona in your whatever county in Arizona you come together and you say "We're going to build a solar co-op." And so then the longer you live there the more invested you are in that solar co-op and you just get a check you just get a check every month from the solar co-op that you're part owner of so that's just an example of those are just a couple of examples of interventions that are possible um employee stock options are another example sovereign wealth funds are other examples there's all kinds of stuff
now you might say okay Dave you proposed a a measurement and some interventions is that really it is that all of post labor economics no we actually when I when I said that we did several years worth of work in an afternoon I am not kidding uh with 03 so what this does is this incentivizes several other downstream things
so the the next measurement that we came up with is what we called Collective Purchasing Power so collective purchasing power is when you look at the purchasing power of a county so the purchasing power of a county is how much capital they can mobilize without increasing taxes or debt and so it's like okay if if you if you have higher purchasing power in a county that means hey we've got a budget basically TLDDR we've got a budget surplus where are we going to invest that budget surplus are we going to build a data center are we going to build a solar farm are we going to build better internet infrastructure how are we going to invest this and so what you want to do is you want to increase your your collective purchasing power over time
um so your CPP will go up over time uh so that's Collective Purchasing Power EAI is your Economic Agency Index and you want your Economic Agency Index to go up over time so CPP and EI are both highly correlated and you want them to go up over time right now these are not being measured by the way we made up these measurements and this is all still hypothetical we need to run the numbers and we need to do some natural experiments but the higher the CPP of a county the more it can reinvest in its own infrastructure and the wealthier it'll get because the more people it will attract um and and humans will be the will be the economic principles they will always be the economic principles of the future
um the Alaska permanent fund uh the their oil fund is another example of uh of what you can do so if a county has some natural resources whether it's farmland or timberland or you know fisheries or whatever you you you socialize it you're not socializing it um we're working on the terminology because we want to avoid all those common pitfalls anyways getting lost in the weeds
so what this does though is it incentivizes different sets of behaviors um so one of the things that it incentivizes is if there is an underutilized county it incentivizes people to go say "Hey this county has a lot of untapped potential let me actually move back out to a rural county where there's not a whole lot of people but there's a lot of there's a lot of unt untapped resources whether it's solar or hydro or fisheries or whatever but because you're managing it locally sustainability is going to be top of mind
if you move to a county that um that has like good estuaries you're going to want to preserve those estuaries because number one those estuaries are going to be productive so you want those biomes to be healthy you want your forest to be healthy so it's not going to be as as extractive so for instance instead of instead of a a company that's just going to come in fish it all out harvest all the logs and leave you are invested locally so by tying your interest to the local land you're going to be far more interested in circular economies in sustainable and renewable technologies and those sorts of things so there you've we've already with this post-labor economic model we've already incentivized stabilizing the environment and preserving the environment because guess what if you harvest everything it stops producing and you stop getting your uh your dividend checks in the mail
so that's number one number two is it it disincentivizes brain drain so basically the longer you stay in a place um the more invested you are in in its success now what I will say though is that we don't want to discourage people from moving and leaving if they want to if there's a better opportunity somewhere else so you'll what this does is that you'll naturally find an equilibrium where there's going to be basically kind of a carrying capacity for each county and so it's like once a county gets too crowded um the the the the distribution of revenues will start to go down so people will be like well you know my my my monthly checks were were 1,500 uh from this county now they're 500 so I'm going to find somewhere better to go so it will create a naturally self-correcting population distribution
so you're going to have some counties that have more carrying capacity due to tech jobs due to natural resources due to other desirable factors uh you know Malibu for instance Los Angeles County it's it's it's going to it's going to it's going to stay up there just because it's desirable and more people are going to be coming in some pe some people are going to get priced out which is fine that happens anyways gentrification is a real thing but instead of saying,"Well I'm going to get pushed out of Los Angeles and now I can't afford anything because there's no jobs." You say "Let me go to an underpopulated county where instead of you know 500 bucks a month from Los Angeles County I get 2500 bucks a month from you know this this rural county." And so then that will that will cause a not just a redistribution of wealth that will cause a redistribution of population which will alleviate lots and lots of other social and economic bottlenecks because cities are very efficient in some respects but they're also very expensive
i could keep going on and on and on for literally hours about this but I think you get the idea so I've still got a lot of work to do to to button up post labor economic theory but it's coming um and long story short if we if we help create counties and and push counties in a direction where you don't need wages but they're still productive and people who live there have ownership stakes and everything we've solved post labor economics you don't need work your work is helping the county manage its resources and develop those resources um and then if you want to move from one county to another you can and we've worked on we've worked out all the schemes for that because like one of the things would be uh would be dividend seekers or or uh or rent tourists dividend tourists so like basically if you say well I'm not making much money from this county so let me just move to another county soak up some dividends and then leave you can do that but you but what you want to do is you want to create some friction so there's some there's a vesting schedule there's the the option to buy in or buy out um your your stake all kinds of things
and by the way none of this none of this is exotic or new all of this is all established the only things that I'm proposing that are new are those are those two metrics there's a third one that I'm not going to talk about yet um so there's three metrics and the idea that that the interventions should be done at the county level but again I didn't come up with subsidiary it's just like this is this is the right principle to use so literally all I'm doing is I'm I'm proposing three metrics and then a set of interventions all the interventions have already been proven out in reality
so yeah post-labor economics is uh hypothetically obviously it hasn't been tested yet um but has hypothetically been solved um I just need to put it together put together a dashboard and start start building a case it could be wrong right one of one of the things that that 03 said is you know here's here's a few questions to ask yourself and one of them is like which of your predictions would have to fail for you to to give up on post labor economics and I'm like that's a good question i don't know all right ramble done thanks for watching let me know what you thought about the format and yeah I'm ultra excited 03 is like feel we we have felt the AGI all right later
