We want to maximize our output, and in this case, for the worse. We also look for shortcuts, and AI provides those shortcuts since writing is difficult and time-consuming for most people. Since we have been given a tool for doing it, we have become dependent on it.
My recommendation, even before Gen AI proliferation, was to read old books (25-30+ years old and, if not 100s years old) as much as possible. The reason is that pressure to produce was not there in the past, and also, if a book or concept has survived for so long and is still applicable, then Lindy’s effect has become relevant to it. Most new ideas, concepts, and books have not gone through the test of time, so in most cases, wait to read them and see if they still exist after 15-20 years of publishing. That does not mean I do not read new ideas or books; I do, but only if I can validate the concepts in other ways. I believe in the Royal Society’s motto, 'Nullius in verba,' which means 'take nobody's word for it' unless I can validate it in another way.
I will end with one of my favorite quotes from C. S. Lewis. He observed that every age suffers from its own blindness — failing to recognize perspectives that will be obvious to succeeding generations. To overcome such blindness, he writes, “The only palliative is to keep the clean breeze of the centuries blowing through our minds, and this can be done only by reading old books.”
I think that very well applies to non fiction writing. So many flash in the pan ideas end up fading a way with a whimper but the author just churns out another with the label of 'best selling.'
On the other hand, how will a book ever be picked up in 30 years of its never read now. That's an authors dilemma.
When I write, my goal is to be timeless. I don't want to be in 'the current thing' even though it does influence my thoughts. However, in 10 years I trust you could go back to my archives and find the material still stuck the landing as it were.
This is valid to a large extent. I also said that I read new books. However, you have to be very selective when reading new books, and that’s what my message above is about. Not every old book we read today was popular in its time, but it became one since it has a timeless value.
You cannot avoid reading the latest books about new technology or innovation as there may not be anything old available, or even if it is available, it may be obsolete.
About fiction, I “mostly” read them for fun rather than to get some new ideas.
Sorry, I wasn't meaning to disagree but to riff off of it. There are a lot of books that weren't successful at the time but then became crazy. However there are a lot of the self help type that fail after just a few years.
It's a good recommendation to look at the older books. I find myself doing the same.
I did not feel like that at all. I do like back-and-forth as it evolves my thinking, too. I know I will not always be correct or cannot think of every possibility. Also, I am a sample size of 1 with my experience, so I know my limitations very well. All this is my attempt to solidify my ideas and then put them out there to see if they are worth keeping or if I need to change.
I want more people to disagree (not just for the sake of) with me at work and in non-work life (my wife does a good job with the latter). I openly encourage disagreement at work.
We need more disagreements so that good ideas prevail, as most people follow the saying, "Just go along to get along."
AI is a tool. Tools are amoral. I would guess from the first tool made, we have used them for good and evil. Will AI bring about our betterment? Our deception? Our destruction? At the individual, community, or species level?
I have been fascinated with this concept since Cory Doctorow coined the word a few years ago. It is amazing how broadly applicable and appealing it is.
Part of the problem with AI-generated content is that is driven by a thoughtless drive to be “visble” saying something……in turn driven by endless push to sell stuff that nobody needs in the first place.
Humans screw up. To really screw things up, you need a computer. To scale from a class A screwup to an apocalyptic disaster, you need to connect computers to the Internet. 30 years ago, we did not have the Internet--a powerful tool that not only scales the best and worse parts of humanity but it provides a platform to people who can barely string words together. I don't blame social media, or AI, or even people. It is the scalability.
This is yet another good example of when understanding ourselves will lead to a greater understanding of the impacts of AI. I think a lot about the idea that pretraining data comes from search a lot of the time, and search comes from AI a lot of the time, but you're right to point out that this has always been the case, just with humans. It's just that the feedback loop is magnified, so it's like the wheel is spinning at 1000 RPM instead of like 10 as before.
It's the amplification that's notable, right? We've always leveraged both our strengths and our faults with whatever has been available... but the tool itself is just much more capable, so it's just the same forces at play, but magnified as never before.
I guess it's the same as it has always been, but also different due to the speed and intensity. Maybe that's my nuanced hot take for the week.
We want to maximize our output, and in this case, for the worse. We also look for shortcuts, and AI provides those shortcuts since writing is difficult and time-consuming for most people. Since we have been given a tool for doing it, we have become dependent on it.
My recommendation, even before Gen AI proliferation, was to read old books (25-30+ years old and, if not 100s years old) as much as possible. The reason is that pressure to produce was not there in the past, and also, if a book or concept has survived for so long and is still applicable, then Lindy’s effect has become relevant to it. Most new ideas, concepts, and books have not gone through the test of time, so in most cases, wait to read them and see if they still exist after 15-20 years of publishing. That does not mean I do not read new ideas or books; I do, but only if I can validate the concepts in other ways. I believe in the Royal Society’s motto, 'Nullius in verba,' which means 'take nobody's word for it' unless I can validate it in another way.
I will end with one of my favorite quotes from C. S. Lewis. He observed that every age suffers from its own blindness — failing to recognize perspectives that will be obvious to succeeding generations. To overcome such blindness, he writes, “The only palliative is to keep the clean breeze of the centuries blowing through our minds, and this can be done only by reading old books.”
I think that very well applies to non fiction writing. So many flash in the pan ideas end up fading a way with a whimper but the author just churns out another with the label of 'best selling.'
On the other hand, how will a book ever be picked up in 30 years of its never read now. That's an authors dilemma.
When I write, my goal is to be timeless. I don't want to be in 'the current thing' even though it does influence my thoughts. However, in 10 years I trust you could go back to my archives and find the material still stuck the landing as it were.
This is valid to a large extent. I also said that I read new books. However, you have to be very selective when reading new books, and that’s what my message above is about. Not every old book we read today was popular in its time, but it became one since it has a timeless value.
You cannot avoid reading the latest books about new technology or innovation as there may not be anything old available, or even if it is available, it may be obsolete.
About fiction, I “mostly” read them for fun rather than to get some new ideas.
Sorry, I wasn't meaning to disagree but to riff off of it. There are a lot of books that weren't successful at the time but then became crazy. However there are a lot of the self help type that fail after just a few years.
It's a good recommendation to look at the older books. I find myself doing the same.
I did not feel like that at all. I do like back-and-forth as it evolves my thinking, too. I know I will not always be correct or cannot think of every possibility. Also, I am a sample size of 1 with my experience, so I know my limitations very well. All this is my attempt to solidify my ideas and then put them out there to see if they are worth keeping or if I need to change.
I want more people to disagree (not just for the sake of) with me at work and in non-work life (my wife does a good job with the latter). I openly encourage disagreement at work.
We need more disagreements so that good ideas prevail, as most people follow the saying, "Just go along to get along."
Valid.
AI is a tool. Tools are amoral. I would guess from the first tool made, we have used them for good and evil. Will AI bring about our betterment? Our deception? Our destruction? At the individual, community, or species level?
That's what I found in writing a lot about AI. It is a tool and can help us learn more about ourselves.
I have been fascinated with this concept since Cory Doctorow coined the word a few years ago. It is amazing how broadly applicable and appealing it is.
And how we've always dealt with it.
Part of the problem with AI-generated content is that is driven by a thoughtless drive to be “visble” saying something……in turn driven by endless push to sell stuff that nobody needs in the first place.
Exactly right. Thankfully it's derivative and easier to see for what it is.
You keep nailing these essays! Keep up the great work!
Humans screw up. To really screw things up, you need a computer. To scale from a class A screwup to an apocalyptic disaster, you need to connect computers to the Internet. 30 years ago, we did not have the Internet--a powerful tool that not only scales the best and worse parts of humanity but it provides a platform to people who can barely string words together. I don't blame social media, or AI, or even people. It is the scalability.
See https://tomrearick.substack.com/p/understanding-ai-risk
That's a great point and actually an under-current for my novel on AI and the apocalypse it unlocks.
https://www.thesingularitychronicles.com
So... What percentage of "bullshit" is the internet filled with?
As primary enshittification or secondary?
😆 ... As cumulative sum. 😁
99%
This is yet another good example of when understanding ourselves will lead to a greater understanding of the impacts of AI. I think a lot about the idea that pretraining data comes from search a lot of the time, and search comes from AI a lot of the time, but you're right to point out that this has always been the case, just with humans. It's just that the feedback loop is magnified, so it's like the wheel is spinning at 1000 RPM instead of like 10 as before.
Exactly. Everytime I look at AI I find terrible human behaviors amplified.
It's the amplification that's notable, right? We've always leveraged both our strengths and our faults with whatever has been available... but the tool itself is just much more capable, so it's just the same forces at play, but magnified as never before.
I guess it's the same as it has always been, but also different due to the speed and intensity. Maybe that's my nuanced hot take for the week.