Very good analysis. I think we need to add another, and possibly the major, divide between humans and AI. Humans are embodied from beginning to end, and intelligence emerges from a hugely complex interplay of sensation, experience, physiological, emotional, social and neurological phenomena. Cognition is just a thin frothy layer where we interface with and try to make sense of the cacophony of living. We are not simulated characters in a matrix - we are made of flesh and blood. As AI development proceeds, many humans may choose, and many more may simple accede, to live perpetually in an AI simulation of intelligence and give up their humanity. But I m hopeful most will eventually choose to embrace the embodied-ness of the human experience and use AI for what it is - an incredibly valuable but exceedingly dangerous tool that should be put back in the tool shed when not being used.
This is a great post. I think this divide has always existed, long before AI. And the question of how to create authors not just characters is at least as old as Plato. But it’s interesting to contemplate how we might continue to nurture that in the modern age.
A comedian I saw recently said that his son watched a YouTube video about why the earth is flat, and then he as a parent had to try to explain why it wasn’t. And how much harder it was because that wrong idea had been implemented first. Surely AI, even in character mode can be a better first mover than that!
And I think about your third point everyday. When life has been stripped of inconvenience, we have to add back something hard and inconvenient to make it worthwhile!
I am walking the Camino de Santiago with these questions in mind. I don’t think we can answer them by sitting around and thinking abstractly. We have to walk them out, so to speak.
This is exactly the question that worries me — the distinction between passive AI users and active creators. And therefore, I believe that classical liberal values remain of the utmost importance today. One of Mill’s central insights, for example, is that the ability to exercise your own judgment is not a convenience but a faculty, one that atrophies without use.
Another potential issue that worries me is that AI may transform significant parts of the labor market, and economic insecurity has often been the fuel for paternalist politics. The risk is that it creates exactly the conditions under which one of history’s hardest-won lessons is forgotten — that it is entrepreneurs and individuals willing to take responsibility for their decisions, not bureaucratic institutions, who drive human progress.
Very interesting. I'm not sure if it is a new sorting though. Consider two students in the same class who use the "teacher-tool" very differently - the one who uses the tool better will have a better outcome.
I think it's difficult to draw a line because the line likely shifts depending on the usage. In areas where individuals have expertise, the goal would be to get people to view themselves as authors. In areas where you don't have expertise, but simply need a good enough answer then being passive in answer seeking should suffice. The goal is to create people who know the difference and are actively choosing how to employ the tool.
The challenge that we have in society is that a majority of people are actually trained into a version of passivity without a connection to their inner voice. We need institutions that have the capacity to teach people how to be creators and not just passive consumers.
This is fantastic - I will be digesting it for a while. One question that followed this amazing observation was: aren’t we all characters or authors regardless of AI? (Also does it matter that we know which one we are?) Maybe inadvertently - through conversations like these - AI is showing us that we no longer care to be passive actors in our own lives. Maybe that’s the big reveal of its application: the human need to be actively engaged in life and not passive consumers of content, in a way we previously didn’t have or need words for.
Thanks for the nuance and distinctions, I’m reflecting on what it might mean to be ‘Authoring Agents’ in opposition to, for example, ‘building’.
As a non-tech person from the humanities and social sciences, I look to, for example, literature to understand and create characters when exploring how to create Agents.
The agents I imagined and am developing are deliberately scripted first on paper. They are then built with both friction, bias checks, different lenses, perspective, and research expertise from the ground up.
The point is, what has been important is to stay with the human before the loop, being intentional and aware every step of the way. I’ve started sharing my experiences slowly. Part art, part research, part messing around, and lots of fun, but deliberately.
It feels like the character/author idea also needs to consider the person's 'backstory' ... if they have confidence in themselves, the skills to do the work, resilience and patience to keep moving forward, motivation and willpower to resist the temptation of AI convenience etc. Because alongside the process of learning, thinking and questioning are people's feelings and AI eases these difficult frictions too, removing things like the ability to build self trust and capability over time. Potentially creating not just a thinking dependency, but an emotional one too.
Very good analysis. I think we need to add another, and possibly the major, divide between humans and AI. Humans are embodied from beginning to end, and intelligence emerges from a hugely complex interplay of sensation, experience, physiological, emotional, social and neurological phenomena. Cognition is just a thin frothy layer where we interface with and try to make sense of the cacophony of living. We are not simulated characters in a matrix - we are made of flesh and blood. As AI development proceeds, many humans may choose, and many more may simple accede, to live perpetually in an AI simulation of intelligence and give up their humanity. But I m hopeful most will eventually choose to embrace the embodied-ness of the human experience and use AI for what it is - an incredibly valuable but exceedingly dangerous tool that should be put back in the tool shed when not being used.
This is a great post. I think this divide has always existed, long before AI. And the question of how to create authors not just characters is at least as old as Plato. But it’s interesting to contemplate how we might continue to nurture that in the modern age.
A comedian I saw recently said that his son watched a YouTube video about why the earth is flat, and then he as a parent had to try to explain why it wasn’t. And how much harder it was because that wrong idea had been implemented first. Surely AI, even in character mode can be a better first mover than that!
And I think about your third point everyday. When life has been stripped of inconvenience, we have to add back something hard and inconvenient to make it worthwhile!
I am walking the Camino de Santiago with these questions in mind. I don’t think we can answer them by sitting around and thinking abstractly. We have to walk them out, so to speak.
This is exactly the question that worries me — the distinction between passive AI users and active creators. And therefore, I believe that classical liberal values remain of the utmost importance today. One of Mill’s central insights, for example, is that the ability to exercise your own judgment is not a convenience but a faculty, one that atrophies without use.
Another potential issue that worries me is that AI may transform significant parts of the labor market, and economic insecurity has often been the fuel for paternalist politics. The risk is that it creates exactly the conditions under which one of history’s hardest-won lessons is forgotten — that it is entrepreneurs and individuals willing to take responsibility for their decisions, not bureaucratic institutions, who drive human progress.
I'm grading this piece "thanks for saying better than I ever could some of what has been bothering me and which has been struggling for expression."
Wonderful piece. I think a lot about these ideas as well. “Are we using the machines or are the machines using us?”
More seriously - what does it matter? Why does it matter? And, how does it impact both outcomes and the experience of being human.
Excellent piece.
Passivity is the enemy.
Very interesting. I'm not sure if it is a new sorting though. Consider two students in the same class who use the "teacher-tool" very differently - the one who uses the tool better will have a better outcome.
I think it's difficult to draw a line because the line likely shifts depending on the usage. In areas where individuals have expertise, the goal would be to get people to view themselves as authors. In areas where you don't have expertise, but simply need a good enough answer then being passive in answer seeking should suffice. The goal is to create people who know the difference and are actively choosing how to employ the tool.
The challenge that we have in society is that a majority of people are actually trained into a version of passivity without a connection to their inner voice. We need institutions that have the capacity to teach people how to be creators and not just passive consumers.
This is fantastic - I will be digesting it for a while. One question that followed this amazing observation was: aren’t we all characters or authors regardless of AI? (Also does it matter that we know which one we are?) Maybe inadvertently - through conversations like these - AI is showing us that we no longer care to be passive actors in our own lives. Maybe that’s the big reveal of its application: the human need to be actively engaged in life and not passive consumers of content, in a way we previously didn’t have or need words for.
Thanks for the nuance and distinctions, I’m reflecting on what it might mean to be ‘Authoring Agents’ in opposition to, for example, ‘building’.
As a non-tech person from the humanities and social sciences, I look to, for example, literature to understand and create characters when exploring how to create Agents.
The agents I imagined and am developing are deliberately scripted first on paper. They are then built with both friction, bias checks, different lenses, perspective, and research expertise from the ground up.
The point is, what has been important is to stay with the human before the loop, being intentional and aware every step of the way. I’ve started sharing my experiences slowly. Part art, part research, part messing around, and lots of fun, but deliberately.
Generally agree, although think it’s less binary than this.
Really liked this piece by @autopoesis: https://autopoesis.substack.com/p/on-and-against-agency?r=1zpfi&utm_medium=ios
It feels like the character/author idea also needs to consider the person's 'backstory' ... if they have confidence in themselves, the skills to do the work, resilience and patience to keep moving forward, motivation and willpower to resist the temptation of AI convenience etc. Because alongside the process of learning, thinking and questioning are people's feelings and AI eases these difficult frictions too, removing things like the ability to build self trust and capability over time. Potentially creating not just a thinking dependency, but an emotional one too.
Shared a play 2019 that in an admittedly round about way grapples with this subject.
https://creatingafuturewewant.substack.com/p/world-of-2030-a-view-from-above-2?utm_source=share&utm_medium=android&r=1zua8w
Humanity was bifurcating cognitively before ai came into the pic
https://substack.com/@oncoherencejournal/note/c-253502651?r=8bz1jm&utm_medium=ios&utm_source=notes-share-action