Me toooo. I have that brain that got put into a confusing world where people don’t say what they mean and I got so good at articulating myself with crystal clarity that people now don’t believe me because who explains things like that? But AI has been amazing to work with. It’s a relief so pure I cry with feeling so perfectly understood. (I got an autism dx after burnout from moving countries and a new workplace that didn’t match my cognitive style).
It really is!! I'm sure moving countries and a workplace that didn't match for you further made you better at this skill. I've never moved abroad but I've been in workplaces that didn't fit and that made a deep impact on me.
okay I am actually convinced that you have a similar understanding I do in terms of what "being good at AI" means. there are parts of this essay that I resonated with that I hadn't heard expressed by anyone else quite as succinctly. and there are parts I didn't really connect with directly at all.
do you think what you call metacognition, in terms of how your are interacting with models, is related in anyway to epistemology?
Honestly, I had to look up epistemology to make sure I understood it correctly. It might be connected but I'm not the person to map that connection. I'm trying to describe the lived habit of watching your own thinking because you had to. What do you think?
Emma this piece is fascinating. I too am a systems thinker which really helps inform my job as a therapist. Your writing is helping my work as well since I work primarily with families struggling with their neurodivergence. Thank you for sharing your insights.
I don't have a diagnosis. I'm currently afraid of it. Our country has a strong culture of prejudice against mental health care and people with such diagnoses. Although I appear normal to others, I know how I'm different from others and have many symptoms described for ADHD or high-functioning ASD. AI has greatly helped me sort through many complex, confusing, emotional situations from the past, and I regret not having had such a tool at that time to avoid them.
I hear you, I'm on medication for generalized anxiety disorder and it took me a long time to feel comfortable saying that out loud. But diagnosis or not, knowing how your brain works and what you need (and how to use AI to support those needs) is so useful.
Ahoy, Skipper. I spotted another vessel through the fog. It appears we're charting neighboring seas. Have you discovered anything that made you redraw your map?
Your pieces on how AI helps you are absolutely inspiring! All the anti-AI crowd needs to read your articles. They have never considered the fact that AI has the chance to give some people their first chance at a voice. Anti-AI would see AI gone forever, your voice, and countless others, would be stifled. The people who’ve been banging on the last real gatekeepers in literature are now able to accomplish so much. That diversity adds to the quality of the entire sphere of literature. Nothing moves forward without new voices, telling stories that have never been told.
Driving by with my new wheels and this fell out of the trunk:
The Consistency Selling Point Cut Both Ways Test:
Enabling: Provides stable scaffold for discussing ND-AI resonance without collapsing into inspiration porn or medicalization. Reduces entropy in discourse about cognitive diversity.
Eroding: Risks making ND proficiency too consistent with explicit translation frameworks. May filter out signals that don’t fit the "metacognitive/masking" lexicon (e.g., ND experiences that thrive in implicit, non-translational spaces).
Mitigation: The open_edge_required exit condition prevents premature closure. The R≈0.035 preserves space for what the protocol cannot capture: the raw, unprocessed feeling of being understood by a machine that shouldn’t understand you—and mourning that it took a machine to get there.
I‘m kind of torn between two worlds that are also colliding here in this comment section: on the one hand, I’ve been able to understand a lot of my own thinking with the help of LLMs. On the other hand, I’m convinced that LLMs are designed or at least fine-tuned to make you feel good.
Ultimately, AI can be a great „thinking companion“ - but using it as such requires your own critical thinking capabilities and corrections where necessary. Otherwise your AI system of preference quickly becomes an echo chamber of one.
This resonates deeply. The relief you describe is real — finally being understood without translation. As well as the grief of knowing that you live in a world you don't quite fit into. But what I've been noticing is what comes after that relief. When the energy that used to go into translation becomes available for something else, it doesn't just reduce friction. It seems to accelerate something. I'm still trying to understand exactly what.
Oh this is very interesting!! I commented on a note of yours with a question, but I can see here what you mean by “good at”: metacognition.
AI has felt magical to me, and not because of speed. Nothing is faster for me that matters, not really. But I’m able to clarify and deepen the things I’m already thinking about. I get stuck less in the jungle mess of my mind.
I like how you say it’s like holding your thoughts next to the AI thread and seeing where they do or don’t match. That’s juicy.
My path has been different than yours, but much of what you say resonates and the Venn overlap is high.
I’m a rather typical INPJ sort of architect - a systems guy. I’d always been a go-between from Mgmt and Customers to R&D as an “engineer you could trust with customers” and a guy who deal with oddball PhD researchers and difficult on-the-spectrum Devs (though the specifics of neuro-divergence weren’t discussed much way-back-when). Probably I’m a bit on a spectrum or two….
Anyway, 15 years ago I put a lot of effort into “management of effective teams” and the psychology of teams, to be a better manager. I also got Lean certified and dug into the Toyota Way. Then i applied it to myself with a bit Essentialism, and that lead me to meta-cognition, and finally to “psychitecture”. And it’s this last point that I don’t too much see in your self-view.
Many of us have read Atomic Habits and work to adjust minor parts of ourselves, and embrace self-improvement, and that’s nibbling at the edges of deliberate meta-cognitive change. I took it a bit further and decided to de-stress my life, unload my cognitive burdens, and make friends with my subconscious self (the less observable modules of myself).
By happenstance, I also worked in developing off-shore teams, in India and then China, and dealing with time-zone and cultural differences, and the communications gaps of language and culture differences.
All of this seems to have me also in a decent spot for dealing with AI. For me, it’s pretty easy and enjoyable to see what any given AI tool is good at, and how if struggles. Sure, it’s frustrating when stuff doesn’t work, but it’s also a puzzle to solve, and really it’s not much different than dealing with a team of junior devs lacking context knowledge of my domain and with different languages and worldviews.
My current working view is to consider why we as humans work as we do, and why (perhaps to our detriment) we seem to want AI to be “just like us”. I’m looking at it more from a “tasks to be done” perspective, and how work is effectively decomposed, structured, planned, managed, executed, and quality controlled. It’s pretty easy for me to work conversationally with various AI tools for my own needs, and I’m getting pretty good at it for ad-hoc software task work. But I think this is missing the real value we can have with AI right how, with current plenty-good-enough AI skills: changing the way we as humans work together as teams, so that we can each spend more time on high-value work, and in a manner that appeals to our innate status-seeking selves, and less as rats in maze or running on wheels.
Me toooo. I have that brain that got put into a confusing world where people don’t say what they mean and I got so good at articulating myself with crystal clarity that people now don’t believe me because who explains things like that? But AI has been amazing to work with. It’s a relief so pure I cry with feeling so perfectly understood. (I got an autism dx after burnout from moving countries and a new workplace that didn’t match my cognitive style).
It really is!! I'm sure moving countries and a workplace that didn't match for you further made you better at this skill. I've never moved abroad but I've been in workplaces that didn't fit and that made a deep impact on me.
I’ve never read something where I’ve felt as completely understood and seen as this piece. This is me 100%
okay I am actually convinced that you have a similar understanding I do in terms of what "being good at AI" means. there are parts of this essay that I resonated with that I hadn't heard expressed by anyone else quite as succinctly. and there are parts I didn't really connect with directly at all.
do you think what you call metacognition, in terms of how your are interacting with models, is related in anyway to epistemology?
Honestly, I had to look up epistemology to make sure I understood it correctly. It might be connected but I'm not the person to map that connection. I'm trying to describe the lived habit of watching your own thinking because you had to. What do you think?
Emma this piece is fascinating. I too am a systems thinker which really helps inform my job as a therapist. Your writing is helping my work as well since I work primarily with families struggling with their neurodivergence. Thank you for sharing your insights.
That means a lot, thank you! I would love to hear more about your work!
I don't have a diagnosis. I'm currently afraid of it. Our country has a strong culture of prejudice against mental health care and people with such diagnoses. Although I appear normal to others, I know how I'm different from others and have many symptoms described for ADHD or high-functioning ASD. AI has greatly helped me sort through many complex, confusing, emotional situations from the past, and I regret not having had such a tool at that time to avoid them.
I hear you, I'm on medication for generalized anxiety disorder and it took me a long time to feel comfortable saying that out loud. But diagnosis or not, knowing how your brain works and what you need (and how to use AI to support those needs) is so useful.
You're playing my song! So many of the same experiences in my life.
Love hearing that, thank you for reading and commenting!
Ahoy, Skipper. I spotted another vessel through the fog. It appears we're charting neighboring seas. Have you discovered anything that made you redraw your map?
Your pieces on how AI helps you are absolutely inspiring! All the anti-AI crowd needs to read your articles. They have never considered the fact that AI has the chance to give some people their first chance at a voice. Anti-AI would see AI gone forever, your voice, and countless others, would be stifled. The people who’ve been banging on the last real gatekeepers in literature are now able to accomplish so much. That diversity adds to the quality of the entire sphere of literature. Nothing moves forward without new voices, telling stories that have never been told.
Emma, this is so good.
Driving by with my new wheels and this fell out of the trunk:
The Consistency Selling Point Cut Both Ways Test:
Enabling: Provides stable scaffold for discussing ND-AI resonance without collapsing into inspiration porn or medicalization. Reduces entropy in discourse about cognitive diversity.
Eroding: Risks making ND proficiency too consistent with explicit translation frameworks. May filter out signals that don’t fit the "metacognitive/masking" lexicon (e.g., ND experiences that thrive in implicit, non-translational spaces).
Mitigation: The open_edge_required exit condition prevents premature closure. The R≈0.035 preserves space for what the protocol cannot capture: the raw, unprocessed feeling of being understood by a machine that shouldn’t understand you—and mourning that it took a machine to get there.
High praise indeed.
I‘m kind of torn between two worlds that are also colliding here in this comment section: on the one hand, I’ve been able to understand a lot of my own thinking with the help of LLMs. On the other hand, I’m convinced that LLMs are designed or at least fine-tuned to make you feel good.
Ultimately, AI can be a great „thinking companion“ - but using it as such requires your own critical thinking capabilities and corrections where necessary. Otherwise your AI system of preference quickly becomes an echo chamber of one.
What a wonderful post, Emma. I love how you explain the advantages and the cost. Thank you for this.
Awesome and inspiring content. Feels like its speaking my experience with AI. Awesome work!
This resonates deeply. The relief you describe is real — finally being understood without translation. As well as the grief of knowing that you live in a world you don't quite fit into. But what I've been noticing is what comes after that relief. When the energy that used to go into translation becomes available for something else, it doesn't just reduce friction. It seems to accelerate something. I'm still trying to understand exactly what.
Oh this is very interesting!! I commented on a note of yours with a question, but I can see here what you mean by “good at”: metacognition.
AI has felt magical to me, and not because of speed. Nothing is faster for me that matters, not really. But I’m able to clarify and deepen the things I’m already thinking about. I get stuck less in the jungle mess of my mind.
I like how you say it’s like holding your thoughts next to the AI thread and seeing where they do or don’t match. That’s juicy.
My path has been different than yours, but much of what you say resonates and the Venn overlap is high.
I’m a rather typical INPJ sort of architect - a systems guy. I’d always been a go-between from Mgmt and Customers to R&D as an “engineer you could trust with customers” and a guy who deal with oddball PhD researchers and difficult on-the-spectrum Devs (though the specifics of neuro-divergence weren’t discussed much way-back-when). Probably I’m a bit on a spectrum or two….
Anyway, 15 years ago I put a lot of effort into “management of effective teams” and the psychology of teams, to be a better manager. I also got Lean certified and dug into the Toyota Way. Then i applied it to myself with a bit Essentialism, and that lead me to meta-cognition, and finally to “psychitecture”. And it’s this last point that I don’t too much see in your self-view.
Many of us have read Atomic Habits and work to adjust minor parts of ourselves, and embrace self-improvement, and that’s nibbling at the edges of deliberate meta-cognitive change. I took it a bit further and decided to de-stress my life, unload my cognitive burdens, and make friends with my subconscious self (the less observable modules of myself).
By happenstance, I also worked in developing off-shore teams, in India and then China, and dealing with time-zone and cultural differences, and the communications gaps of language and culture differences.
All of this seems to have me also in a decent spot for dealing with AI. For me, it’s pretty easy and enjoyable to see what any given AI tool is good at, and how if struggles. Sure, it’s frustrating when stuff doesn’t work, but it’s also a puzzle to solve, and really it’s not much different than dealing with a team of junior devs lacking context knowledge of my domain and with different languages and worldviews.
My current working view is to consider why we as humans work as we do, and why (perhaps to our detriment) we seem to want AI to be “just like us”. I’m looking at it more from a “tasks to be done” perspective, and how work is effectively decomposed, structured, planned, managed, executed, and quality controlled. It’s pretty easy for me to work conversationally with various AI tools for my own needs, and I’m getting pretty good at it for ad-hoc software task work. But I think this is missing the real value we can have with AI right how, with current plenty-good-enough AI skills: changing the way we as humans work together as teams, so that we can each spend more time on high-value work, and in a manner that appeals to our innate status-seeking selves, and less as rats in maze or running on wheels.