Mark Palmer - long time since StreamBase! - one thing I've learned recently working with engineering adopting AI-powered engineering is that the skeptics become some of the most powerful users.
In the case of engineers, there is the camp that is "vibe-coding" and just happy something mostly-works. They've adopted AI quickly to get them going. There is another camp that is disciplined, professional, lives & breathes details, and has tended to be AI skeptic (hallucinations? mistakes? more to cleanup!).
However these AI skeptics tend to be the most powerful users in practice -- they write prompts that are precise, they carefully craft their workflows to get the most out of the AI tools. They just have to shift their thinking and how they leverage AI.
Mark! So great to hear from you. This is why I love to write. And, wow, that is so well said -- indeed, often, the biggest skeptics turn out to be the biggest fans. I remember a few great StreamBase customers who were wary of visual development that turned out to be huge fans, after digging in.
The issue I've had is with a different kind of skeptic: "Seagull AI Skeptics." These are people who fly around and shit all over everything, without giving it a sincere try. These are engaged even less than your group of "Happy to just something going."
Then, on top of that, there are the "AI Ethnographers." These are your group that turn prompting and working with AI output into a craft. They carefully understand how LLMs behave and "live and breathe in the details," as you say.
The group that I wrote this about are the Seagull Skeptics. The folks who simply shit all over AI for, in many cases, good theoretical reasons, but simply won't try to use it. That's the tough group, because to understand AI, you must work with AI. Until you engage, worries about bias, hallucinations, and ethics reign; once you engage, they melt away, or at least you become part of solving the issues, not just complaining about them.
"They just have to shift their thinking and how they leverage AI" -- that's kind of the battle I'm fighting -- to help folks shift their thinking. But it's hard!
I’m leaning toward optimism, Mark. :) The “AI steals” perspectives tends to strike me as more cautious than constructive. No one really knows where AI is headed long-term, but I’m seeing the benefits in my day-to-day, and I’m thankful that thoughtful guardrails are being built as we go. Different opinions will help the tech grow in the right direction—with the right checks in place.
My thought at the moment: Humans have always built on existing work whether in art, music, literature, or science. Literature review/background study in research, for example, was all about that. What we used to do with a pen and index cards (!!!) was replaced by internet search. Now, AI takes that even further. I’m grateful to have a creative tool that amplifies our ability to explore and synthesize what already exists—while still leaving room for uniquely human insight.
Hi Sachi -- It's so interesting you mention Noah Smith. I just got through listening to him on the Big Technology Podcast. He certainly is a skeptic of another kind -- in this episode, calling BS on the incredible 10s-of-billions market caps of AI companies: https://open.spotify.com/episode/5wdvOribedaT9FNTv59OIj?si=Y84mZydJSB2efaqK3p0n9Q
I like the article you shared by him better; it's basically saying we shouldn't pretend to know where this will all lead. Agree! But that's the fun of it, too.
I, like you, LEAN toward "Love." But like Noah, cringe at the over-hypers, too. His cartoon of "humans need not apply" is one of my most hated memes (broad proclamations of the demise of human work. In fact, my continued favorite point of view on human work is the one economist David Autor from MIT shared years ago -- his "Never Get Enough" principle, that humans constantly find new things to do with each technology wave. Although his talk is pre-AI-craze, I still think it strikes the perfect balance of embracing the fear of automation and skill replacement with the evolution and optimism of the future of work:
Hi Mark, thank you for sharing these resources. I came across the O-Ring and Neger Get Enough theories for the first time. They were both eye-opening, brilliant analyses! David Autor’s point that “material abundance has never eliminated perceived scarcity” truly captures how far we’ve come as humans, yet how much we still feel is missing.
That idea resonates with how I experience AI in my own work. Tasks I once did manually, like synthesizing information, can now be done much faster with AI. But the work doesn’t end there; it creates space to build on that foundation and add more value through uniquely human insight and creativity. It’s like building a second brain—not to replace thinking, but to enhance it. So again, I’m optimistic and excited about what AI can bring.
I really enjoyed your Maven session a few years ago (so inspiring!) and continue to look forward to your future articles!
Hi Mark, I wanted to share Morgan Housel's recent podcast, which reinforces David Autor’s “Never Get Enough” theory (though not by name)! He noted how automation has historically created entirely new industries and roles—60% of U.S. jobs in 2018 didn’t exist in 1940. My takeaway is that while negative feelings are real and shouldn’t be minimized, history shows that society has faced similar upheavals before—and we’ve evolved! Adapting to rapid technological change seems a timeless challenge. https://podcasts.apple.com/us/podcast/what-ai-might-do-to-the-future-of-work/id1675310669?i=1000721104729
Are we making tools or idols? for the fear is that we have become addicted to the quick fix, the "life hack", and bowing to the false god of "efficiency". Endlessly searching, hunting for the magic algorithm that will bring maximum profit with minimum effort
Yes. Sadly, research shows that 68% (MIT Study) to 80% (McKinsey study) of AI users use it as a quick fix, life hack, or a magic algorithm. In all of this writing, the theme is about how to avoid the "magical thinking" fallacy and "using it with vigor."
I think this is the bigger issue beyond potential bias, energy ethics, or creative rights -- that AI triggers humans to be lazy. That concerns me more.
This may trigger a bunch of negative blowback, because many do not want to or like to consider God in all these matters, but here's something I heard clearly in 2022:
Here's the link to the Outlier conference that has the replay of the talk; I think you need to pay to watch the replay, unfortunately. That said, there were some amazing talks there that are worth paying for from leading data visualization experts like the lead data graphics person from the New York Times and experts from all backgrounds. Truly inspirational and good stuff. https://www.outlierconf.com/
It’s on the Outlier website, but it requires registration to view it. I’ll post that link in a minute. That said, the gist of my talk can be found in this blog post, which I wrote in the month preceding the event as a way to sort out my thoughts. The article, ironically, is much more aggressive about the threat of AI on the field of data analysis.
Thanks Amanda. I don’t mind challenge. The irony is that this talk wasn’t even about the more interesting topics among the many fascinating areas of AI. This one was pretty basic. But so much of the “discourse” wasn’t constructive.
Indeed, it’s time to move on! Thanks for commenting.
I'm in the same boat and I see the same things. I even have an essay coming up on the plagarism of AI or, at least, the accusation of that. When you hear words like theft, and plagarism, I just have to accept that they don't know what they're talking about and it's a bad look becasue it means they don't really understand the work they do.
Thanks Mike. Honestly, it's nice to read your comment because this post was really, mostly, a purging of the frustrations I've encountered with these debates. The AI theft topic is a terrific one, as I find myself over and over asking those who argue for theft for specifics, citing the New York Times lawsuit with OpenAI where the NYT's lawyers clearly manipulated their prompts to get ChatGPT to regurgitate Times content (that came from other stolen sources on the internet.). Here it is, again, here: https://openai.com/index/openai-and-journalism/
Anyway, when I find myself shadow boxing with myself, it's clear that I'm wasting my time, although I've also had some really good, productive conversations as well. Thx for the comment!
It's so frustrating and so it's nice to see that I'm not the only one. It doesn't help that folks like Gary Marcus and Philip Koopman keep pouring gas on the fire with one outrage after another. I just wonder, at a certain point, isn't the constant reaction exhausting?
Mark Palmer - long time since StreamBase! - one thing I've learned recently working with engineering adopting AI-powered engineering is that the skeptics become some of the most powerful users.
In the case of engineers, there is the camp that is "vibe-coding" and just happy something mostly-works. They've adopted AI quickly to get them going. There is another camp that is disciplined, professional, lives & breathes details, and has tended to be AI skeptic (hallucinations? mistakes? more to cleanup!).
However these AI skeptics tend to be the most powerful users in practice -- they write prompts that are precise, they carefully craft their workflows to get the most out of the AI tools. They just have to shift their thinking and how they leverage AI.
Mark! So great to hear from you. This is why I love to write. And, wow, that is so well said -- indeed, often, the biggest skeptics turn out to be the biggest fans. I remember a few great StreamBase customers who were wary of visual development that turned out to be huge fans, after digging in.
The issue I've had is with a different kind of skeptic: "Seagull AI Skeptics." These are people who fly around and shit all over everything, without giving it a sincere try. These are engaged even less than your group of "Happy to just something going."
Then, on top of that, there are the "AI Ethnographers." These are your group that turn prompting and working with AI output into a craft. They carefully understand how LLMs behave and "live and breathe in the details," as you say.
The group that I wrote this about are the Seagull Skeptics. The folks who simply shit all over AI for, in many cases, good theoretical reasons, but simply won't try to use it. That's the tough group, because to understand AI, you must work with AI. Until you engage, worries about bias, hallucinations, and ethics reign; once you engage, they melt away, or at least you become part of solving the issues, not just complaining about them.
"They just have to shift their thinking and how they leverage AI" -- that's kind of the battle I'm fighting -- to help folks shift their thinking. But it's hard!
I’m leaning toward optimism, Mark. :) The “AI steals” perspectives tends to strike me as more cautious than constructive. No one really knows where AI is headed long-term, but I’m seeing the benefits in my day-to-day, and I’m thankful that thoughtful guardrails are being built as we go. Different opinions will help the tech grow in the right direction—with the right checks in place.
My thought at the moment: Humans have always built on existing work whether in art, music, literature, or science. Literature review/background study in research, for example, was all about that. What we used to do with a pen and index cards (!!!) was replaced by internet search. Now, AI takes that even further. I’m grateful to have a creative tool that amplifies our ability to explore and synthesize what already exists—while still leaving room for uniquely human insight.
Slightly different context, but I recently came across this thoughtful analysis on AI pessimism from economist Noah Smith, which resonated with me: https://www.noahpinion.blog/p/stop-pretending-you-know-what-ai
Hi Sachi -- It's so interesting you mention Noah Smith. I just got through listening to him on the Big Technology Podcast. He certainly is a skeptic of another kind -- in this episode, calling BS on the incredible 10s-of-billions market caps of AI companies: https://open.spotify.com/episode/5wdvOribedaT9FNTv59OIj?si=Y84mZydJSB2efaqK3p0n9Q
I like the article you shared by him better; it's basically saying we shouldn't pretend to know where this will all lead. Agree! But that's the fun of it, too.
I, like you, LEAN toward "Love." But like Noah, cringe at the over-hypers, too. His cartoon of "humans need not apply" is one of my most hated memes (broad proclamations of the demise of human work. In fact, my continued favorite point of view on human work is the one economist David Autor from MIT shared years ago -- his "Never Get Enough" principle, that humans constantly find new things to do with each technology wave. Although his talk is pre-AI-craze, I still think it strikes the perfect balance of embracing the fear of automation and skill replacement with the evolution and optimism of the future of work:
https://www.ted.com/talks/david_autor_will_automation_take_away_all_our_jobs
Hi Mark, thank you for sharing these resources. I came across the O-Ring and Neger Get Enough theories for the first time. They were both eye-opening, brilliant analyses! David Autor’s point that “material abundance has never eliminated perceived scarcity” truly captures how far we’ve come as humans, yet how much we still feel is missing.
That idea resonates with how I experience AI in my own work. Tasks I once did manually, like synthesizing information, can now be done much faster with AI. But the work doesn’t end there; it creates space to build on that foundation and add more value through uniquely human insight and creativity. It’s like building a second brain—not to replace thinking, but to enhance it. So again, I’m optimistic and excited about what AI can bring.
I really enjoyed your Maven session a few years ago (so inspiring!) and continue to look forward to your future articles!
Hi Mark, I wanted to share Morgan Housel's recent podcast, which reinforces David Autor’s “Never Get Enough” theory (though not by name)! He noted how automation has historically created entirely new industries and roles—60% of U.S. jobs in 2018 didn’t exist in 1940. My takeaway is that while negative feelings are real and shouldn’t be minimized, history shows that society has faced similar upheavals before—and we’ve evolved! Adapting to rapid technological change seems a timeless challenge. https://podcasts.apple.com/us/podcast/what-ai-might-do-to-the-future-of-work/id1675310669?i=1000721104729
Are we making tools or idols? for the fear is that we have become addicted to the quick fix, the "life hack", and bowing to the false god of "efficiency". Endlessly searching, hunting for the magic algorithm that will bring maximum profit with minimum effort
Yes. Sadly, research shows that 68% (MIT Study) to 80% (McKinsey study) of AI users use it as a quick fix, life hack, or a magic algorithm. In all of this writing, the theme is about how to avoid the "magical thinking" fallacy and "using it with vigor."
I think this is the bigger issue beyond potential bias, energy ethics, or creative rights -- that AI triggers humans to be lazy. That concerns me more.
This may trigger a bunch of negative blowback, because many do not want to or like to consider God in all these matters, but here's something I heard clearly in 2022:
https://listenmychild.com/2022/11/25/deception-of-the-digital/
where can I see the video from this event, Mark?
Here's the link to the Outlier conference that has the replay of the talk; I think you need to pay to watch the replay, unfortunately. That said, there were some amazing talks there that are worth paying for from leading data visualization experts like the lead data graphics person from the New York Times and experts from all backgrounds. Truly inspirational and good stuff. https://www.outlierconf.com/
It’s on the Outlier website, but it requires registration to view it. I’ll post that link in a minute. That said, the gist of my talk can be found in this blog post, which I wrote in the month preceding the event as a way to sort out my thoughts. The article, ironically, is much more aggressive about the threat of AI on the field of data analysis.
Here’s the link to the essence of what I said: https://open.substack.com/pub/technosapien/p/rip-bi-the-death-and-rebirth-of-thinking?r=3su05&utm_campaign=post&utm_medium=web&showWelcomeOnShare=false
I liked your talk at Outlier but when I shared my positive thoughts on the group char def got some push back. I think letting it be is the way to go.
Thanks Amanda. I don’t mind challenge. The irony is that this talk wasn’t even about the more interesting topics among the many fascinating areas of AI. This one was pretty basic. But so much of the “discourse” wasn’t constructive.
Indeed, it’s time to move on! Thanks for commenting.
I'm in the same boat and I see the same things. I even have an essay coming up on the plagarism of AI or, at least, the accusation of that. When you hear words like theft, and plagarism, I just have to accept that they don't know what they're talking about and it's a bad look becasue it means they don't really understand the work they do.
Thanks Mike. Honestly, it's nice to read your comment because this post was really, mostly, a purging of the frustrations I've encountered with these debates. The AI theft topic is a terrific one, as I find myself over and over asking those who argue for theft for specifics, citing the New York Times lawsuit with OpenAI where the NYT's lawyers clearly manipulated their prompts to get ChatGPT to regurgitate Times content (that came from other stolen sources on the internet.). Here it is, again, here: https://openai.com/index/openai-and-journalism/
Anyway, when I find myself shadow boxing with myself, it's clear that I'm wasting my time, although I've also had some really good, productive conversations as well. Thx for the comment!
It's so frustrating and so it's nice to see that I'm not the only one. It doesn't help that folks like Gary Marcus and Philip Koopman keep pouring gas on the fire with one outrage after another. I just wonder, at a certain point, isn't the constant reaction exhausting?