Today, for the first time, my Gmail offered to write a response to a piece of mail I got asking for a good time to call me. I shuddered at the thought of it. AI is here to stay and I'm glad I'm old and that I know how to total a column of numbers and be confident of the sum.
Is it possible that the reason we have been unable to detect intelligent life in the universe is that all advanced life forms inevitably create an artificial intelligence, which then ... (your Ai-pocalypse here)
I can’t decide what it means yet either, but I’m comforted by the fact I’ll be either a vegetable in a nursing home (on an AI assisted monitor no doubt) or dead before the worst happens. Sorry kids, this is your fight. An AI chatbot agreed with me on this, btw, just based on my age.
Have you watched The AI Doc: Or How I Became an Apocaloptimist yet? If not, please do so and then let’s discuss. Before the doc I was worried, then not so worried. Feels like the beginning phase of social media and the internet in the dot com era… well, we know how those things turned out… After watching the doc, now I’m just pissed off that a small number of greedy assholes obsessed with power and money are going to end us. And just about at the point of nothing fucking matters, sell it all, and go travel before this dumbass species wipes itself out… sigh.
I am in the ‘ai is the devil’ category. But really enjoyed these perspectives! Thank you! My personal perspective is: Put the damn technology in box for a decade & let us all potter along as usual. Open the box after a decade. If it’s still useful, put guardrails around it & move forward. (A girl can dream…)
“Then there are those who assure us that this frenetic AI bubble will eventually burst. That there is not enough information in existence for these models to scrape in order to achieve next-level usefulness. That, eventually, what will happen is that the training models will rely more and more on their own slop, thus baking into the models all the inaccuracies, hallucinations, and flattened-thinking that limits current models. In other words, they will, eventually, eat their own tails.”
This is something I have been thinking about and it makes sense when the chatbot is trained on user queries and creative works are created with the machine doing most of the content generation. Then if everyone wants to be Mr. Beast or PewDiePie, I think the urge to create content too quickly exacerbates the attention span problem we already have and also amplifies the hyperbolistic and provocative attention seeking with often a polemical tone or else callous comedy.
We gotta linger more on each idea and ask ourselves more questions. I’m haunted by what I learned in psychology classes about neuroplasticity— the parts of our brain that we use frequently are strengthened and those we don’t use are not developed. Also, the hippocampus and medial temporal lobe supposedly thin or shrink if we are sedentary all the time, leading to more difficulty with learning and memory formation. Horrifying for mobility impaired disabled people. I guess I have to separate out the sedentary lifestyle from the passive receipt of artificial intelligence generated content. But I keep connecting these experiences and imagining the early onset of dementia.
We need to be more interactive with the content we receive and expand upon it. But I suspect most of us are lazy thinkers who will get lazier.
Getting from here to the magical world of everyone having a universal basic income because The Macines can do it all.. Hold on to your hat. And I wouldn't stop saving for retirement just yet...
I'm a software engineer and this is how fast it's moving: none of our software engineers write their own code anymore. That is over. Claude Code writes better, faster, cleaner, more secure code. But no engineers have been laid off. We're super busy figuring out all the new stuff we can do with our AI and it's a lot! Our jobs, so far, are just changing -- and it's pretty exciting. That said, I still feel lucky I'm retiring in 3 months.
Sonja, really? Same here - software engineer, but Claude is not writing better code than us. In fact, I was explicitly hired to fix Claude's code. I don't know how long this will last, but at the moment, Claude isn't totally there in my opinion. However I agree that the job definition is changing, and it's becoming more about telling the computer what to do than how to do it. You would think that this would mean all product and design people can now become engineers, but no - it's been interesting to see how not everyone is up for the task of thinking through every possible edge case in the logic. I wouldn't say I'm even up for it, but historically this is how I've gotten paid, so I continue to do it. Jealous that you're retiring.
Ask Claude to fix it. Then ask Claude to test it. Are you building Claude agents? Then get multiple Claude agents running and testing each other’s code. Have Claude check the security, any glitches, syntax, anything. Have Claude document everything. It will remember how to fix things long after the one developer who remembered is retired.
A few months ago I googled to find an article I’d once read wherein Rob Burnett related the origin of the Godfrey running gag on “Ed” (too much time on my hands? Probably.), and the instant AI response referenced Arthur Godfrey as a cast member, I believe playing Phil! It was the most convoluted slop, I wish I’d saved it because it was hilarious, but was mostly corrected when I tried a second time.
I know there will be good and bad in AI, but it feels ominous to me. Godfrey agrees.
“I don’t know what to believe.” To a certain extent I share this agnosticism 🤷♂️. But one critical question, it seems to me, is whether there is a merely accidental or contingent correlation between AI and the techno-fascist project of the broligarchy, or whether it bespeaks a necessary and inevitable connection. If it turns out to be, on compelling analysis, the latter, I would say that my model for the wise use of technological advancement would be Captain Nemo.
At this point, after reading a lot and using LLMs for different tasks, I have come to think of it as a demon-esq kind of entity. It's not my friend, it's not on my side, it won't necessarily tell the truth all the time. But it's quite useful. Show it a shelf of bottles at the Asian grocery and it translates all of it so I can buy the right sauce...handy.
A human (me!) wrote the words to this song, but AI wrote the music and IS the voices. It is a great tool for creating protest songs with cumbersome titles!
Today, for the first time, my Gmail offered to write a response to a piece of mail I got asking for a good time to call me. I shuddered at the thought of it. AI is here to stay and I'm glad I'm old and that I know how to total a column of numbers and be confident of the sum.
Is it possible that the reason we have been unable to detect intelligent life in the universe is that all advanced life forms inevitably create an artificial intelligence, which then ... (your Ai-pocalypse here)
I can’t decide what it means yet either, but I’m comforted by the fact I’ll be either a vegetable in a nursing home (on an AI assisted monitor no doubt) or dead before the worst happens. Sorry kids, this is your fight. An AI chatbot agreed with me on this, btw, just based on my age.
Have you watched The AI Doc: Or How I Became an Apocaloptimist yet? If not, please do so and then let’s discuss. Before the doc I was worried, then not so worried. Feels like the beginning phase of social media and the internet in the dot com era… well, we know how those things turned out… After watching the doc, now I’m just pissed off that a small number of greedy assholes obsessed with power and money are going to end us. And just about at the point of nothing fucking matters, sell it all, and go travel before this dumbass species wipes itself out… sigh.
I do notice less typos in your posts in the last few months. Jus' sayin'
I am in the ‘ai is the devil’ category. But really enjoyed these perspectives! Thank you! My personal perspective is: Put the damn technology in box for a decade & let us all potter along as usual. Open the box after a decade. If it’s still useful, put guardrails around it & move forward. (A girl can dream…)
Well-written and thoughtful, as always.
“Then there are those who assure us that this frenetic AI bubble will eventually burst. That there is not enough information in existence for these models to scrape in order to achieve next-level usefulness. That, eventually, what will happen is that the training models will rely more and more on their own slop, thus baking into the models all the inaccuracies, hallucinations, and flattened-thinking that limits current models. In other words, they will, eventually, eat their own tails.”
This is something I have been thinking about and it makes sense when the chatbot is trained on user queries and creative works are created with the machine doing most of the content generation. Then if everyone wants to be Mr. Beast or PewDiePie, I think the urge to create content too quickly exacerbates the attention span problem we already have and also amplifies the hyperbolistic and provocative attention seeking with often a polemical tone or else callous comedy.
We gotta linger more on each idea and ask ourselves more questions. I’m haunted by what I learned in psychology classes about neuroplasticity— the parts of our brain that we use frequently are strengthened and those we don’t use are not developed. Also, the hippocampus and medial temporal lobe supposedly thin or shrink if we are sedentary all the time, leading to more difficulty with learning and memory formation. Horrifying for mobility impaired disabled people. I guess I have to separate out the sedentary lifestyle from the passive receipt of artificial intelligence generated content. But I keep connecting these experiences and imagining the early onset of dementia.
We need to be more interactive with the content we receive and expand upon it. But I suspect most of us are lazy thinkers who will get lazier.
Getting from here to the magical world of everyone having a universal basic income because The Macines can do it all.. Hold on to your hat. And I wouldn't stop saving for retirement just yet...
I'm a software engineer and this is how fast it's moving: none of our software engineers write their own code anymore. That is over. Claude Code writes better, faster, cleaner, more secure code. But no engineers have been laid off. We're super busy figuring out all the new stuff we can do with our AI and it's a lot! Our jobs, so far, are just changing -- and it's pretty exciting. That said, I still feel lucky I'm retiring in 3 months.
Sonja, really? Same here - software engineer, but Claude is not writing better code than us. In fact, I was explicitly hired to fix Claude's code. I don't know how long this will last, but at the moment, Claude isn't totally there in my opinion. However I agree that the job definition is changing, and it's becoming more about telling the computer what to do than how to do it. You would think that this would mean all product and design people can now become engineers, but no - it's been interesting to see how not everyone is up for the task of thinking through every possible edge case in the logic. I wouldn't say I'm even up for it, but historically this is how I've gotten paid, so I continue to do it. Jealous that you're retiring.
Ask Claude to fix it. Then ask Claude to test it. Are you building Claude agents? Then get multiple Claude agents running and testing each other’s code. Have Claude check the security, any glitches, syntax, anything. Have Claude document everything. It will remember how to fix things long after the one developer who remembered is retired.
I regret that your probably well written, insightful and absorbing article arrived this Monday morning.
Give me until tomorrow to read anything about AI (A-eye). No can do today.
Well I know I’M human, but we had a goodish run.
I’m old enough to remember when it was mandatory for robots to obey the 3 Laws. There were anomalies but it mostly kept us alive.
A few months ago I googled to find an article I’d once read wherein Rob Burnett related the origin of the Godfrey running gag on “Ed” (too much time on my hands? Probably.), and the instant AI response referenced Arthur Godfrey as a cast member, I believe playing Phil! It was the most convoluted slop, I wish I’d saved it because it was hilarious, but was mostly corrected when I tried a second time.
I know there will be good and bad in AI, but it feels ominous to me. Godfrey agrees.
“I don’t know what to believe.” To a certain extent I share this agnosticism 🤷♂️. But one critical question, it seems to me, is whether there is a merely accidental or contingent correlation between AI and the techno-fascist project of the broligarchy, or whether it bespeaks a necessary and inevitable connection. If it turns out to be, on compelling analysis, the latter, I would say that my model for the wise use of technological advancement would be Captain Nemo.
At this point, after reading a lot and using LLMs for different tasks, I have come to think of it as a demon-esq kind of entity. It's not my friend, it's not on my side, it won't necessarily tell the truth all the time. But it's quite useful. Show it a shelf of bottles at the Asian grocery and it translates all of it so I can buy the right sauce...handy.
A human (me!) wrote the words to this song, but AI wrote the music and IS the voices. It is a great tool for creating protest songs with cumbersome titles!
https://ohhisteve.substack.com/p/three-fellas-in-search-of-pardons?r=3b8a3e&utm_medium=ios