AI may radically change our conception of “who is in an org”.
Historically, there is a sharp delineation between an org’s staff (on the “inside”), versus its users (on the “outside”). The relationship of a fulltime employee of an org is radically different than that of one of its customers. This is especially true for consumer tech startups & companies: think Google, Meta, Reddit, X, now perhaps OpenAI and Anthropic.
With AI however, the bar for an external user to be able to contribute, drops drastically. At the same time, employing and compensating this user will be much easier, as AI obliterates transaction costs. Overall, I expect org boundaries to become much more porous: more part-time workers, more volunteers, more collectives and communes.
Stepping back: why is it the case that orgs have a relative few fulltime employees? What makes it hard for a single org to stretch to encompass more humans? I’ll go through four major historical drivers, and how AI might change each.
First, it’s hard to hire and onboard. Hiring requires a lot of trust. An org wants to know that a person is a good fit, worth paying for. The job market is famously a market for lemons. Orgs try to vet candidates, but resumes are low-signal; interviews cost precious staff time, and also select for people good at interviewing. Ideally you would see the person doing the job — in a work test or work trial.
AI can change this by making search and matching problems much more efficient. And this is not just “pair up existing resumes & job listings”. AI can form a much more detailed model of a human based on all their public outputs, and a role’s requirements based on all of its outputs. Or: have your Claude apply for jobs, have their Claude interview your Claude to see if the fit is good.
Second, larger teams incur higher communication costs. The number of pairwise relationships grows quadratically with team size; Dunbar famously pointed to a cap of ~150 people per stable community. Startup literature preaches the gospel of small nimble teams.
Prosaic technology (Slack, Google Calendar, Notion, Zoom) is already enabling much larger & profitable orgs. AI can improve on this in several ways: connecting data sources smoothly (eg MCP); making it easier to spin up custom software to serve internal use cases, and (speculatively) by simulating other humans, letting one speak with a colleague’s digital twin.
Third, on-the-job learning historically leads to specialization. As an employee continues in an org, they build up skills and tacit knowledge that are hard to replace, sometimes unique to their particular role. You’ll notice that great orgs work hard at staff retention. However, this kind of specialization is a double-edge sword for the employee. While it helps with job security, it also means that tenured employees accrue all kinds of responsibilities which hold back new work.
As more work is intermediated via LLM, LLM systems can hopefully take on more of specialization, first via in-context learning, then building up memory to reinforce particular skillsets. Humans today are still better at generalizing from few-shot examples, so their comparative advantage may remain as generalists, attacking new domains.
Fourth: humans are just really expensive! Paying for staff compensation is a major check on the growth of orgs. Successful orgs today draw an economic boundary around a few extremely great people, and pay them handsomely in cash and equity.
But in the long run AIs will drive down the cost of human upkeep. Surplus and abundance driven by rapid technical growth means that human consumption needs can be met much more easily. When people no longer rely on a paycheck to cover needs, they can spend their time with orgs which help fulfill their higher values.
Given all this, how might orgs operate in the future?
Maybe: membership & employment norms would change. You might work 5 “jobs” each at 1-10h/week. You might switch between orgs the way you hop between different social media sites today, polyamorously, without jealousy.
Maybe: you might switch between jobs much faster. If transaction costs are low, if you can onboard and offboard faster, your median tenure at a job might change from 2 years to 2 months (or, 2 weeks), as you impart the essential wisdom and craft that make you “you” — then parachute off to the next opportunity.
Maybe: we might end up with vast, nation-level superorgs, like the hives of Terra Ignota, entities glued together with AI support and radically different governance. Imagine X, except if the money was universal, people drew significant paychecks from their posts and work, and there’s taxation, election representation rather than just top-down Elon-controlled. (What is a network state in the era of transformative AI?)
Why am I excited for these more porous orgs? We’ve seen how solid org boundaries can lead to an “us vs them” mentality. Enshittification or shortform videos or gambling addictions or doomy content sucking out unendorsed hours from its prey. And, at the extreme end, xrisk: frontier lab employees endangering the rest of humanity.
If instead such an org could wholly integrate its leaders, staff and users, it might be unstoppable. I admire OpenAI’s commitment to offering free intelligence to the world; Anthropic’s commitment to spending its gains for the public benefit. Imagine if these orgs were not just serving, but also governed by all of humanity, by virtue of having all of humanity on the inside!
Thanks to @Lydia Nottingham for reviewing an early outline. This was intended as a submission to an essay competition, though I may not have submitted in time.
AI may radically change our conception of “who is in an org”.
Historically, there is a sharp delineation between an org’s staff (on the “inside”), versus its users (on the “outside”). The relationship of a fulltime employee of an org is radically different than that of one of its customers. This is especially true for consumer tech startups & companies: think Google, Meta, Reddit, X, now perhaps OpenAI and Anthropic.
With AI however, the bar for an external user to be able to contribute, drops drastically. At the same time, employing and compensating this user will be much easier, as AI obliterates transaction costs. Overall, I expect org boundaries to become much more porous: more part-time workers, more volunteers, more collectives and communes.
Stepping back: why is it the case that orgs have a relative few fulltime employees? What makes it hard for a single org to stretch to encompass more humans? I’ll go through four major historical drivers, and how AI might change each.
First, it’s hard to hire and onboard. Hiring requires a lot of trust. An org wants to know that a person is a good fit, worth paying for. The job market is famously a market for lemons. Orgs try to vet candidates, but resumes are low-signal; interviews cost precious staff time, and also select for people good at interviewing. Ideally you would see the person doing the job — in a work test or work trial.
AI can change this by making search and matching problems much more efficient. And this is not just “pair up existing resumes & job listings”. AI can form a much more detailed model of a human based on all their public outputs, and a role’s requirements based on all of its outputs. Or: have your Claude apply for jobs, have their Claude interview your Claude to see if the fit is good.
Second, larger teams incur higher communication costs. The number of pairwise relationships grows quadratically with team size; Dunbar famously pointed to a cap of ~150 people per stable community. Startup literature preaches the gospel of small nimble teams.
Prosaic technology (Slack, Google Calendar, Notion, Zoom) is already enabling much larger & profitable orgs. AI can improve on this in several ways: connecting data sources smoothly (eg MCP); making it easier to spin up custom software to serve internal use cases, and (speculatively) by simulating other humans, letting one speak with a colleague’s digital twin.
Third, on-the-job learning historically leads to specialization. As an employee continues in an org, they build up skills and tacit knowledge that are hard to replace, sometimes unique to their particular role. You’ll notice that great orgs work hard at staff retention. However, this kind of specialization is a double-edge sword for the employee. While it helps with job security, it also means that tenured employees accrue all kinds of responsibilities which hold back new work.
As more work is intermediated via LLM, LLM systems can hopefully take on more of specialization, first via in-context learning, then building up memory to reinforce particular skillsets. Humans today are still better at generalizing from few-shot examples, so their comparative advantage may remain as generalists, attacking new domains.
Fourth: humans are just really expensive! Paying for staff compensation is a major check on the growth of orgs. Successful orgs today draw an economic boundary around a few extremely great people, and pay them handsomely in cash and equity.
But in the long run AIs will drive down the cost of human upkeep. Surplus and abundance driven by rapid technical growth means that human consumption needs can be met much more easily. When people no longer rely on a paycheck to cover needs, they can spend their time with orgs which help fulfill their higher values.
Given all this, how might orgs operate in the future?
Maybe: membership & employment norms would change. You might work 5 “jobs” each at 1-10h/week. You might switch between orgs the way you hop between different social media sites today, polyamorously, without jealousy.
Maybe: you might switch between jobs much faster. If transaction costs are low, if you can onboard and offboard faster, your median tenure at a job might change from 2 years to 2 months (or, 2 weeks), as you impart the essential wisdom and craft that make you “you” — then parachute off to the next opportunity.
Maybe: we might end up with vast, nation-level superorgs, like the hives of Terra Ignota, entities glued together with AI support and radically different governance. Imagine X, except if the money was universal, people drew significant paychecks from their posts and work, and there’s taxation, election representation rather than just top-down Elon-controlled. (What is a network state in the era of transformative AI?)
Why am I excited for these more porous orgs? We’ve seen how solid org boundaries can lead to an “us vs them” mentality. Enshittification or shortform videos or gambling addictions or doomy content sucking out unendorsed hours from its prey. And, at the extreme end, xrisk: frontier lab employees endangering the rest of humanity.
If instead such an org could wholly integrate its leaders, staff and users, it might be unstoppable. I admire OpenAI’s commitment to offering free intelligence to the world; Anthropic’s commitment to spending its gains for the public benefit. Imagine if these orgs were not just serving, but also governed by all of humanity, by virtue of having all of humanity on the inside!
Thanks to @Lydia Nottingham for reviewing an early outline. This was intended as a submission to an essay competition, though I may not have submitted in time.