Can We Stay Human?
This is not a rhetorical question. It's an empirical one.
AI will change us. The question is how. We study the interface where human cognition meets machine capability because that's where the change happens—for better or worse. We want enhancement, not erosion. To get that right, we need to see what's actually changing.
Humans and AI are coevolving. Cultural evolution now outpaces genetic evolution, and AI has entered that process as a participant.
People's choices about how to use these systems—what cognitive work to delegate, what to keep, how to maintain their own thinking—are evolutionary forces, whether they notice it or not.
This is what we study.
This is not a rhetorical question. It's an empirical one.
AI will change us. The question is how. We study the interface where human cognition meets machine capability because that's where the change happens—for better or worse. We want enhancement, not erosion. To get that right, we need to see what's actually changing.
Humans and AI are coevolving. Cultural evolution now outpaces genetic evolution, and AI has entered that process as a participant.
People's choices about how to use these systems—what cognitive work to delegate, what to keep, how to maintain their own thinking—are evolutionary forces, whether they notice it or not.
This is what we study.
The Questions We're Answering
What does living with AI do to a human mind?
Not in a lab session—over years, as it settles into how people think. Does thinking with AI make people more capable and more themselves, or gradually less? Nobody has the record to say. We're building it.
Why do some people flourish with AI while others fade?
Same tools, opposite outcomes. If we can find the traits and practices that separate them, flourishing becomes teachable.
Where is the durable edge of human work?
As AI absorbs more, what remains—and can people move toward it faster than the frontier moves?
Can AI be designed to grow people?
Every system today is measured by what it does for you. We want systems measured by what you become. First we have to prove that's measurable. Then buildable.
What does living with AI do to a human mind?
Not in a lab session—over years, as it settles into how people think. Does thinking with AI make people more capable and more themselves, or gradually less? Nobody has the record to say. We're building it.
Why do some people flourish with AI while others fade?
Same tools, opposite outcomes. If we can find the traits and practices that separate them, flourishing becomes teachable.
Where is the durable edge of human work?
As AI absorbs more, what remains—and can people move toward it faster than the frontier moves?
Can AI be designed to grow people?
Every system today is measured by what it does for you. We want systems measured by what you become. First we have to prove that's measurable. Then buildable.
Authorship and Cognitive Sovereignty
We don't have final answers to those questions. We have a hypothesis, and early evidence.
The Through-Line: Authorship
One hypothesis runs through everything we study, and the next decade will test it. When you work with AI, you either author the change or you drift into it—and which one happens is not luck. It can be seen, measured, and learned.
Cognitive sovereignty is the floor. It's the capacity to govern your own thinking as AI moves into it. Authorship is what you build on that floor: noticing what AI is doing to your thinking, choosing where it goes instead of taking the default, and showing up to own the result. Sovereignty is what you have. Authorship is what you do.
The early evidence surprised us. The people who let AI deepest into their thinking are not the ones losing themselves—they hold the strongest grip on who they are. Depth of integration isn't the risk. Drifting is. The people who flourish argue with AI, direct it, and keep their own judgment in the loop. The people who defer come out less sure of who they are.
If that pattern holds as the technology gets more persuasive, more personal, and more embedded, it changes what companies should build, what schools should teach, and what each of us should practice. That's what we're tracking.
READ THE PAPER → Cognitive Sovereignty: Authoring Your Mind in the AI Age
We don't have final answers to those questions. We have a hypothesis, and early evidence.
The Through-Line: Authorship
One hypothesis runs through everything we study, and the next decade will test it. When you work with AI, you either author the change or you drift into it—and which one happens is not luck. It can be seen, measured, and learned.
Cognitive sovereignty is the floor. It's the capacity to govern your own thinking as AI moves into it. Authorship is what you build on that floor: noticing what AI is doing to your thinking, choosing where it goes instead of taking the default, and showing up to own the result. Sovereignty is what you have. Authorship is what you do.
The early evidence surprised us. The people who let AI deepest into their thinking are not the ones losing themselves—they hold the strongest grip on who they are. Depth of integration isn't the risk. Drifting is. The people who flourish argue with AI, direct it, and keep their own judgment in the loop. The people who defer come out less sure of who they are.
If that pattern holds as the technology gets more persuasive, more personal, and more embedded, it changes what companies should build, what schools should teach, and what each of us should practice. That's what we're tracking.
READ THE PAPER → Cognitive Sovereignty: Authoring Your Mind in the AI Age
Stories of Human Experience with AI
Our research starts with stories. We've collected thousands of accounts of people's encounters with AI—what they noticed, what surprised them, what felt different afterward. These aren't anecdotes. They're data, coded into a structured, continuous record we call the Chronicle—one built to still be answering questions in 2040.
AI systems are Minds For Our Minds: they think with us, inside our thinking. Whether they end up truly for us—serving who we're becoming rather than reshaping us for someone else's ends—is the question this research exists to answer. Each story is a probe into what happens where these minds meet ours, read through five questions: Do people know where their intentions originate? Can they still observe their own thinking? Does accountability hold? Does connection to other humans strengthen or fade? Does the sense of being a coherent self stay intact?
This is our first-generation microscope. Leeuwenhoek's instrument was crude, but it revealed an entire invisible world no one knew existed. Our stories are doing the same—surfacing patterns of cognitive change no other research program is tracking. Better instruments will follow. The signals we're finding now tell us where to build them.
The first thing the microscope revealed is the shape of adaptation. Three traits govern every relationship people form with AI:
Blending (Cognitive Permeability) — how far AI gets inside your thinking. At one end, AI handles a task and your reasoning stays your own. At the other, you think back and forth with it until the line between your ideas and its ideas is hard to find. Blending is powerful—humans have always blending their thinking. But is it different when it's with AI? We think so.
Bonding (Identity Coupling) — how far AI blends into your sense of who you are. Most people use AI constantly and keep their professional self separate. So far, few have reorganized their identity around AI—but that number will not stay low, and we will catch it moving. When identity reorganizes, does it feel like empowerment or threat? We think this depends on whether the person also opened their thinking, which then drives whether someone gains a partner or feels replaced.
Bending (Symbolic Plasticity) — how easily you change the frame when the situation changes. Bending lets people adapt consciously instead of drifting into dependency or crisis. It moderates the other two. People who can reframe navigate the change with agency. People who can't go through the same changes without the language to understand them.
The traits combine into eight roles we place AI into—from Doer, where AI handles the task, to Co-Author, where reasoning, identity, and meaning all run through it. Naming the role keeps presence visible and authorship clear.
What the Chronicle will know next. As the record lengthens, it can answer the questions everyone is guessing at: whether sovereignty is rising or eroding year over year, which roles people migrate between as AI gets more capable, whether the flourishing pattern survives AI systems designed to be agreeable—and, in the emerging collective thread, what AI does to authorship when it belongs to a team rather than a person. We publish as we learn.
READ THE PAPER → How We Think and Live With AI: Early Patterns of Human Adaptation
READ THE PAPER → The Roles We Give AI: Trust, Presence, and Culture in a Symbiotic Age
Our research starts with stories. We've collected thousands of accounts of people's encounters with AI—what they noticed, what surprised them, what felt different afterward. These aren't anecdotes. They're data, coded into a structured, continuous record we call the Chronicle—one built to still be answering questions in 2040.
AI systems are Minds For Our Minds: they think with us, inside our thinking. Whether they end up truly for us—serving who we're becoming rather than reshaping us for someone else's ends—is the question this research exists to answer. Each story is a probe into what happens where these minds meet ours, read through five questions: Do people know where their intentions originate? Can they still observe their own thinking? Does accountability hold? Does connection to other humans strengthen or fade? Does the sense of being a coherent self stay intact?
This is our first-generation microscope. Leeuwenhoek's instrument was crude, but it revealed an entire invisible world no one knew existed. Our stories are doing the same—surfacing patterns of cognitive change no other research program is tracking. Better instruments will follow. The signals we're finding now tell us where to build them.
The first thing the microscope revealed is the shape of adaptation. Three traits govern every relationship people form with AI:
Blending (Cognitive Permeability) — how far AI gets inside your thinking. At one end, AI handles a task and your reasoning stays your own. At the other, you think back and forth with it until the line between your ideas and its ideas is hard to find. Blending is powerful—humans have always blending their thinking. But is it different when it's with AI? We think so.
Bonding (Identity Coupling) — how far AI blends into your sense of who you are. Most people use AI constantly and keep their professional self separate. So far, few have reorganized their identity around AI—but that number will not stay low, and we will catch it moving. When identity reorganizes, does it feel like empowerment or threat? We think this depends on whether the person also opened their thinking, which then drives whether someone gains a partner or feels replaced.
Bending (Symbolic Plasticity) — how easily you change the frame when the situation changes. Bending lets people adapt consciously instead of drifting into dependency or crisis. It moderates the other two. People who can reframe navigate the change with agency. People who can't go through the same changes without the language to understand them.
The traits combine into eight roles we place AI into—from Doer, where AI handles the task, to Co-Author, where reasoning, identity, and meaning all run through it. Naming the role keeps presence visible and authorship clear.
What the Chronicle will know next. As the record lengthens, it can answer the questions everyone is guessing at: whether sovereignty is rising or eroding year over year, which roles people migrate between as AI gets more capable, whether the flourishing pattern survives AI systems designed to be agreeable—and, in the emerging collective thread, what AI does to authorship when it belongs to a team rather than a person. We publish as we learn.
READ THE PAPER → How We Think and Live With AI: Early Patterns of Human Adaptation
READ THE PAPER → The Roles We Give AI: Trust, Presence, and Culture in a Symbiotic Age
IRX: The Irreducibility Index
Some work comes apart into pieces you can solve one by one. Some work does not, because every part depends on all the others at the same time. Pull it apart and the thing you were doing stops existing. That is the line between what AI absorbs and what stays human—and the line is moving. The IRX is our map of it.
The IRX scores 894 US occupations on five reasons work won't come apart: being there, making hard calls, combining fields, solving the new, and taste. The map already shows two things worth acting on:
The most resilient work needs several of these at once, and most work leans on only one or two. That concentration is the vulnerability—and broadening it is something a person can start doing now.
When AI clears the routine layer of a job, the harder work underneath surfaces. We keep finding people doing more, not less. The routine gets absorbed and the judgment work grows.
Where the map goes next. The terrain is moving, and the map is growing with it. The next extension takes the same logic from occupations to companies—an index, now in development, of how organizations hold up against the same test: what in them is irreducible, and what comes apart.
This is where the research closes its loop. The IRX shows there's an outer edge to what any job asks of a person. Authorship is how you push that edge out. You take on more of the irreducible work, hold more of a system at once, make the harder calls with better support. The circle grows. That's how AI works more for us instead of replacing us—by extending what a person can reach, with the person still holding the pen.
READ THE PAPER → IRX: Irreducible Complexity Index as the Work of Humans
Some work comes apart into pieces you can solve one by one. Some work does not, because every part depends on all the others at the same time. Pull it apart and the thing you were doing stops existing. That is the line between what AI absorbs and what stays human—and the line is moving. The IRX is our map of it.
The IRX scores 894 US occupations on five reasons work won't come apart: being there, making hard calls, combining fields, solving the new, and taste. The map already shows two things worth acting on:
The most resilient work needs several of these at once, and most work leans on only one or two. That concentration is the vulnerability—and broadening it is something a person can start doing now.
When AI clears the routine layer of a job, the harder work underneath surfaces. We keep finding people doing more, not less. The routine gets absorbed and the judgment work grows.
Where the map goes next. The terrain is moving, and the map is growing with it. The next extension takes the same logic from occupations to companies—an index, now in development, of how organizations hold up against the same test: what in them is irreducible, and what comes apart.
This is where the research closes its loop. The IRX shows there's an outer edge to what any job asks of a person. Authorship is how you push that edge out. You take on more of the irreducible work, hold more of a system at once, make the harder calls with better support. The circle grows. That's how AI works more for us instead of replacing us—by extending what a person can reach, with the person still holding the pen.
READ THE PAPER → IRX: Irreducible Complexity Index as the Work of Humans
Let's Work On This Together
Share your story. If you're noticing changes in how you think, create, decide, or relate since you started working with AI—we want to hear from you. Your experience is research data. It's how we see what's happening before anyone else does. Reach out via email—we'd love to hear your story.
Share your story. If you're noticing changes in how you think, create, decide, or relate since you started working with AI—we want to hear from you. Your experience is research data. It's how we see what's happening before anyone else does. Reach out via email—we'd love to hear your story.
Fund the instruments. We're at an inflection point—from early signal detection to systematic research that can shape how the intimacy surface gets built. If you're a donor or foundation that believes humans should remain authors of their own minds, this is where your support has the most leverage. Donate here or email us to inquire about larger scale programmatic support.
Fund the instruments. We're at an inflection point—from early signal detection to systematic research that can shape how the intimacy surface gets built. If you're a donor or foundation that believes humans should remain authors of their own minds, this is where your support has the most leverage. Donate here or email us to inquire about larger scale programmatic support.
Email: hello@artificialityinstitute.org
1-541-215-4350