This was your best post ever. I thoroughly enjoyed reading it. I actually remember those bank books. I had one as a kid for my savings account. I remember going in with my parents and getting it stamped.
I don't know where everything is going to end up, but things are changing quickly. I agree with you that AI is not the solution to every problem that some people are trying to say that it is. There does need to be caution when using it. It irritates me how people try to sell it as this magical fix.
Thanks for the comments Bette. I thought this was quite a self indulgent post but the range of differing comments meant that I clearly struck a chord. I long thought about doing one of these every two weeks or so and I’m beginning to think that I might want to change to that approach. Who knows?
I got a lot of “we don’t believe in recommendation engines either” kinds of feedback. Do you find that any of them are useful to you?
Wow, I didn’t think it was self-indulgent at all. Honestly, adding those personal anecdotes made it even more relatable, which is what I really enjoyed.
As for your question, no, I don’t find most recommendation engines very useful, and here’s why: If it’s shopping-related, just because I bought something once doesn’t mean I want to keep buying variations of it. For example, on Amazon, buy a sweeper and you probably won’t need another one anytime soon, so the recommendations tend to be a big miss.
Sometimes movie or TV show recommendations get it right. But out of 10 suggestions, I might only be interested in one or two. So, if you think about it that way, the success rate is more like 10%~20%.
Thanks for your comments. I am gradually shifting my focus to a longer biweekly newsletter with a longer opinion piece followed just by stories about AI & business. Fits better with my other newsletter and AI’s future is dependent on business uptake not individuals.
I didn’t get many yay-sayers to recommendation engines.
I believe our world is literally dominated by AI in a good way. I always love the Apple book recommendations and I always love the Fitness check. I also enjoy the daily Screen time control.... as we just keep going we keep seeing how far it takes us David.
Also, I think the retirement plan for the radio house on the island is a good one. That sounds best💯
Not all AI is good, it's invasive and we're abusing its use. There's others that aid with work and we're giving them absolute control when hg shouldn't be so. We turned its original intention and use into what fits us.
Oof, dah, wow, this post resonated on a lot of levels with me.
As far as Ai doing Good Work, a great deal of that is in the sciences where it's used to pick out subtle data patterns that escape human analysis. Or to explore large complex problem spaces for viable solutions (e.g. protein folding). The success seems due in large part to the tight focus of the Ai's goals — in contrast to the general "ask any question" wide focus of popular LLMs. The tightly focused work in the sciences has seriously elevated the game there. That said, I do worry a bit about the *correctness* of even these models. We know LLMs hallucinate badly; I worry that even tightly focused science Ai models might as well.
As far as seeing people fear then embrace technology, in the early 1980s I was a service technician for a large corporation, and a big part of my job was installing fax machines for a variety of corporate customers who were adding faxing to their business process. The orders came from on high, so usually the people actually in the office were facing change and new-fangled technology they didn't understand. They didn't want me there, or my little machine. Six months later they were calling for me to come back desperate because their fax machine wasn't working. Like vampires, once you invite them in, they're hard to get rid of.
Is it too late for us? Yes, I think it is. Something will rise from our ashes, but this civilization is headed for a huge crash, ironically because of both our astonishing species success and our crushing self-defeating stupidity. We're doing a "Tower of Babel" number on ourselves. We've become so mentally decadent that we're wallowing in epistemic chaos. We no longer agree on what's true or ever what truth means (or that it has value). We no longer all operate under the same normative rules established over thousands of years of human history. It's a garden overrun with weeds and run riot. The only hope I see is to plow it under and start fresh.
Back in 2020, on my other blog, I wrote a post, "Our BS Culture", that was about how we've been sliding deeper and deeper into fantasy bullshit (FBS). That epistemic chaos comes from losing touch with physical reality and demonstrable fact. Media technology has played a major role in that. Personal computing turns out to effectively be Huxley's infamous soma, Marx's opiate of the masses. News today is as much about what someone *said* as what actually happened.
Ai recommendations, yeah, ha. I'm someone without terribly strong preferences in what I pick, but I still wouldn't take advice from an algorithm. Ever. I don't quite understand why anyone takes them seriously. To me they have the same low value as commercials. Which is what I think they are. As you point out, they're often pushing product rather than tuning into your preferences. I learned to ignore commercials ages ago; this is just more of the same corporate noise.
One thing I will say is that recommendation requires analysis of your own patterns, of course, but also good curation of the meta-data describing the products. (And, obviously, honesty in the intent to match your preferences with their products.) Good curation usually requires humans with at least some degree of training and reliability, and *that* requires money. I've come to suspect that lack of good meta-data curation is part of the problem.
I also suspect that humanity is too noisy to be a good source of training for these engines because we're all over the map on just about everything. If the Ai is generalizing from that data, enough people with weird preferences may distort the recommendations.
There was a reason why it took me so long to get back to your comments; because I agree with most of them completely. Any AI based on language learning models will have the same structural difficulties so their output also should be checked. There are some people experimenting with a different type of training based on children but I think it will get more human dependent Increasingly some neurosymbolic approaches are cropping up at the edges.
I agree with you that it is too late and for the reasons you cite. Months ago I was chatting with Neela about doom and she thought we agreed. Her version was some bad tough years. Mine was real doom civilization collapse.
I here what you say about the recommendations engines and that perhaps human patterns are difficult to interpret. But in the scientific area they recognize very complicated patterns. I think the commerce motive tied with nostalgia humans overvaluing convenience is why they are successful
I did get a couple of comments that I cannot understand about decision fatigue. I hear people saying there are more decisions to make (which I don’t believe) and live is more stressful. I agree on the work / vocation issue it is. I think the real issue is right in front of them: an extra 3 plus hours dedicated each and every day to scrolling, with a million data points hitting your brain is tiring that what we used to do, which balanced complete downtime with the sustained focus of an activity like reading, listening to music, et al. These aren’t fatiguing activities; scrolling is
I do think we're beyond pendulum swings and see little hope for the immediate future. I think there's a strong element of boiling frog here. Things have changed slowly enough — and modern life has so many kinds of distracting soma — that people just don't realize that not only has the train left the station, but it's a damn way down the line by now.
I wonder if the science applications of Ai are more successful because those applications involve "dumb" matter acting according to physics. With human culture there is the extra component of intelligence (such as it is). I've long thought it ironic that what we call the "soft" sciences are actually incredibly challenging because of the preponderance of variables and human will. Many aspects of physics are extremely difficult, but we make progress in them whereas many aspects of humanity remain stubbornly opaque to us.
But, yeah, I think commercial motives are at least as big of a problem with recommendation engines as the difficultly of parsing human behavior.
Regarding "decision fatigue", I'm surprised you don't believe modern life is faced with vastly more decisions required of us than in the past. TV in the USA used to be three channels broadcasting a fixed schedule. Then there were local channels. And UHF channels. Then cable and now streaming. Movie choices have also expanded from a few theaters to megaplexes to VHS and DVDs and now to streaming platforms. We pick when and what to watch with few restrictions. We have access to an incredible amount of reading material, much of it free. Books, magazines, websites. Not to mention all the social media one can get into. Or consider all the shopping options we have today (e.g. an entire aisle devoted to breakfast cereal). We're bombarded constantly with people trying to sell us their product or get our attention. Myriad causes demand our support (and donations). Even just here on Substack, there are many blogs to choose from.
I quite agree we exhaust ourselves with mindlessness, but I think it's also true that the choices available to us now are dizzying.
I agree wholeheartedly with your frog boiling metaphor.
I think of the concept of complexity as applied to certain types of multivariate situations in terns of causality. So human behavior is one set, but the underlying nature and causes of the many diseases we know as cancer is another. As is climate despite the media’s gross oversimplification around “carbon”
I hear what you say about decisions (vs choice). For us, we have seen the increase in the numbers for some things which should increase our decision fatigue. But it shouldn’t for those borne into it; they should be able to develop strategies to deal with it. Like I have: for most consumer products, say toothpaste and mouthwash, there is no qualitative difference other than do I like the taste. So Find one and keep buying it. But in some cases there is less choice. Like restaurant meals. Three examples. Every pub in North America basically has the same 20 items on the menu, and when one makes a change (to ranch dressing they all follow). There are few creative, exciting options. Second, in every main shopping street in Europe or North America you have the same 25 multinational branded stores. This makes decision making much simpler and they are identikit websites for your digital shopping pleasure. No more unique shops offering some really different product options. Third example, when I was in my prime working years living in Vancouver you could get Ukrainian, Hungarian, Czech, Portuguese, Scandinavian, and 6 types of Chinese. No longer. The restaurant choices are greatly reduced.
If older people have decision fatigue maybe I understand it. But I think the cause in younger people is something deeper and more pernicious.
I don't at all doubt the influence of "something deeper and more pernicious". I'm not sure what to think about younger people who're born into this culture. Definitely there are strategies to deal with it, but even needing those strategies is a resource drain. A higher-energy way of life. Perhaps it's simpler to say that modern life is more exhausting. I just read someone point out how we combine activities now: listen to music while cooking, listen to books while driving, or talking on the cellphone while walking. Too many distractions to ever slow down and just be. So many things to pack into a day.
Maybe it has to do with where we live, but my experience of places to eat is variety. Even "burger joints" offer different sides, kinds of burgers, beers, and vibes. There still are unique shops, though a lot have gone out of business due to Big Box stores. FWIW, I suspect that as 3D printing becomes more of an on-demand shopping option, we may see more variety in products.
Oddly, I'm not sure the lack of variety in culture (which I agree there is truth to — all the chain stores and franchises and movie and book sequels) detracts from there also being so many choices. Even when some channels lack variety, there are still a lot of channels.
The little door was never a problem when it led to fantasy conjured up by books or other art. Somehow that only enhances reality or makes it more bearable. But with technology, we tend to confuse the two states. Why is that?
You are right. But the metaphorical door in technology is profoundly different. I really admit to not understanding why people want to overvalue techie experiences . Just saw another study that found that 45% of under 25s were online continuously
I would say that yes, the majority of us have decision fatigue and want to be told what to do. If that weren't the case, never mind tech, COACHES, wouldn't have a business model!
I realise that when most of us go to a coach they try to guide us to realise the idea ourselves, but a lot of us contract those sorts of services to be told what to do. Damon Mitchell wrote about this recently but I can't tag in comments so I'll send this to him 😊
All this being said, I reckon we all assume that these tech recs come from human input, at some point... We are ignorant of how much is actually machine learning and AI, perhaps deliberately so.
I am interested in the issue of decision fatigue. Which decisions and why the fatigue? If it is around career / vocation / job nexus, I can understand as this part of our life is different and more difficult than in past generations.
All of AI is some statistically moderated version of human information
Personally, it's at work having to say yes and no and maybe to everything my team asks me, also to take in suggestions from other teams and decide which to listen to. At home, then, just deciding what to eat becomes a chore. People who wear and eat the same every day male perfect sense to me, as a result. They make a handful fewer decisions than I do every day 🫡
Thanks for sharing about this. It has been 25 years since I was in the daily operational firing line where decisions were required constantly. I menu planned on Saturday then the decisions were made at a down time. But i love cooking so it helped me get out of my head when I came home.
I could not menu plan to save my life at this point 😆 one day I would love to have the mental bandwidth as well as skill to do that. I rotate very simple meals in my home, partly she to lack of creativity but mostly lack of ability 🍳 any suggestions for a Level 1 Menu Planner?
I laughed at your comment as my wife always says I should write a cookbook for exactly that: healthy, tasty meals that are ready in 10 / 15 / 20 minutes because 2 of our 3 kids and their partners are in panic every dinnertimr. I have a lot of them. Maybe I should.
Thanks for your input. Thus article generated lots of different perspectives. I want to think a out what you said and will het back to you again as I wanted to ask some questions. But now my granddaughter’s demanding fun on Canada Day
Thanks, Nadine. (I'm also unable to tag you.) I'll not share my essay here, lest I seem to be attempting a hijack of the audience. Smart people can find their way to my Substack if they feel so motivated.
As I read this essay, I thought, "Yeah. These engines of choice have never worked that great."
It's ironic that in this world, where GenAI seems to be speeding us along to every valuable conclusion, we still can't help folks figure out which content is available that they will actually want to consume. Just yesterday I started a ChatGPT convo to aggregate my preferences and three movie/series review sites to see if I could hack this more locally. The jury remains out, and I remain hopeful and skeptical.
But I'm not sure I want my GenAI creating more efficiency. Every time I chase efficiency, I have to trade out the wisdom of experience to get it. This is why I don't own a commuter mug or a car to put it in. It's why I don't walk and text, and generally downthrottle every intuition that multitasking is somehow better. It's why I live in Latin America, where "mañana" doesn't mean tomorrow. It means "Not today, Gringo."
My hope is that the backlash to all of this is a kind of slowing down. I'm sure it won't be, but I still hope for that.
Thanks for the comments. I read your latest article. I found your crowfunding metaphor for Chatty GPT refreshingly naive. It may have ingested tomes or Epictetus and the Dalai Lama but also tens of thousands of uninsightful nobodies. It’s massive statistical engine generates primarily mediocre and at times average outputs across a incredibly broad spectrum. I have found with some experiments on subjects that I’m an expert in that AI is an unreliable Encyclopedia with absolutely no real understanding of priorities or preferences. And AI, as currently constituted as LLM‘s are architecturally and conceptually unable to get any better. It has no way of moving toward truth, so despite the underlying technology being known as machine learning, it’s really just doing statistical association of words. The other thing that I find that it does which you mentioned in your peace is that actually disrupt human connection. To me that is a profound negative implication of this technology. However, the conflation of so many other technologies are mean that it’s just one more disruption on the dehumanisation of technology.
Honored that you read my essay. Thank you, David. I'll take "refreshingly naive" as a compliment. At fifty, "bitter and curmudgeonly" are closing in faster than I'd like. Naive, hold me tight! 😉
You're absolutely right, but we all know fast food is not very nutritious, and yet here we are. Even if current models are aggregating bad intel, it's cheap, and it's convenient. I don't think anyone cares or knows enough to care, whatever you and I believe they should do.
Thanks for your reply. Curmudgeonly is a cheap ad hominem attack especially when applied to oneself. I think bunches of individuals using AI for free are incidental to where the technology goes, from an economic standpoint. If it doesn’t work successfully for business organizations (and the real data shows an excess of problems with few successes )the size of the investment is unsustainable. If I could sell OpenAI short I would.
You mention special purpose AI doing real good, but it's hidden behind the hype of general purpose LLMs. This feels like a microcosm of the larger issue - appropriate, measured technological solutions exist, but they can't compete for attention against the flashy alternatives. It's not just that we've become technophiles, but also we've become addicted to technological spectacle. Maybe the solution is better media literacy and a cultural change. Access to information does not equal understanding.
Thanks for your input on these issues. Not sure what will happen going forward. There is a strange fascination with tech spectacle which is the cousin of celebrity worship. We have the growing problem of tinpot ChatGPT experts who do one prompt and bingo!
Sad. It is the never learn that DRIVES ME CRAZY. Great song from the 80s. We never talk to others. We never look back at history. We continually reinvent ourselves like cigarette smokers from the 1950s when we knew smoking killed……..each individual : but not me
Thanks for the restack
This was your best post ever. I thoroughly enjoyed reading it. I actually remember those bank books. I had one as a kid for my savings account. I remember going in with my parents and getting it stamped.
I don't know where everything is going to end up, but things are changing quickly. I agree with you that AI is not the solution to every problem that some people are trying to say that it is. There does need to be caution when using it. It irritates me how people try to sell it as this magical fix.
Thanks for the comments Bette. I thought this was quite a self indulgent post but the range of differing comments meant that I clearly struck a chord. I long thought about doing one of these every two weeks or so and I’m beginning to think that I might want to change to that approach. Who knows?
I got a lot of “we don’t believe in recommendation engines either” kinds of feedback. Do you find that any of them are useful to you?
Wow, I didn’t think it was self-indulgent at all. Honestly, adding those personal anecdotes made it even more relatable, which is what I really enjoyed.
As for your question, no, I don’t find most recommendation engines very useful, and here’s why: If it’s shopping-related, just because I bought something once doesn’t mean I want to keep buying variations of it. For example, on Amazon, buy a sweeper and you probably won’t need another one anytime soon, so the recommendations tend to be a big miss.
Sometimes movie or TV show recommendations get it right. But out of 10 suggestions, I might only be interested in one or two. So, if you think about it that way, the success rate is more like 10%~20%.
Thanks for your comments. I am gradually shifting my focus to a longer biweekly newsletter with a longer opinion piece followed just by stories about AI & business. Fits better with my other newsletter and AI’s future is dependent on business uptake not individuals.
I didn’t get many yay-sayers to recommendation engines.
I think most people find the recommendations from whatever site you're on to be more irritating than helpful.
I believe our world is literally dominated by AI in a good way. I always love the Apple book recommendations and I always love the Fitness check. I also enjoy the daily Screen time control.... as we just keep going we keep seeing how far it takes us David.
Also, I think the retirement plan for the radio house on the island is a good one. That sounds best💯
Thanks for your comments Ral. Bit surprised you think all AI is good
Not all AI is good, it's invasive and we're abusing its use. There's others that aid with work and we're giving them absolute control when hg shouldn't be so. We turned its original intention and use into what fits us.
Oof, dah, wow, this post resonated on a lot of levels with me.
As far as Ai doing Good Work, a great deal of that is in the sciences where it's used to pick out subtle data patterns that escape human analysis. Or to explore large complex problem spaces for viable solutions (e.g. protein folding). The success seems due in large part to the tight focus of the Ai's goals — in contrast to the general "ask any question" wide focus of popular LLMs. The tightly focused work in the sciences has seriously elevated the game there. That said, I do worry a bit about the *correctness* of even these models. We know LLMs hallucinate badly; I worry that even tightly focused science Ai models might as well.
As far as seeing people fear then embrace technology, in the early 1980s I was a service technician for a large corporation, and a big part of my job was installing fax machines for a variety of corporate customers who were adding faxing to their business process. The orders came from on high, so usually the people actually in the office were facing change and new-fangled technology they didn't understand. They didn't want me there, or my little machine. Six months later they were calling for me to come back desperate because their fax machine wasn't working. Like vampires, once you invite them in, they're hard to get rid of.
Is it too late for us? Yes, I think it is. Something will rise from our ashes, but this civilization is headed for a huge crash, ironically because of both our astonishing species success and our crushing self-defeating stupidity. We're doing a "Tower of Babel" number on ourselves. We've become so mentally decadent that we're wallowing in epistemic chaos. We no longer agree on what's true or ever what truth means (or that it has value). We no longer all operate under the same normative rules established over thousands of years of human history. It's a garden overrun with weeds and run riot. The only hope I see is to plow it under and start fresh.
Back in 2020, on my other blog, I wrote a post, "Our BS Culture", that was about how we've been sliding deeper and deeper into fantasy bullshit (FBS). That epistemic chaos comes from losing touch with physical reality and demonstrable fact. Media technology has played a major role in that. Personal computing turns out to effectively be Huxley's infamous soma, Marx's opiate of the masses. News today is as much about what someone *said* as what actually happened.
Ai recommendations, yeah, ha. I'm someone without terribly strong preferences in what I pick, but I still wouldn't take advice from an algorithm. Ever. I don't quite understand why anyone takes them seriously. To me they have the same low value as commercials. Which is what I think they are. As you point out, they're often pushing product rather than tuning into your preferences. I learned to ignore commercials ages ago; this is just more of the same corporate noise.
One thing I will say is that recommendation requires analysis of your own patterns, of course, but also good curation of the meta-data describing the products. (And, obviously, honesty in the intent to match your preferences with their products.) Good curation usually requires humans with at least some degree of training and reliability, and *that* requires money. I've come to suspect that lack of good meta-data curation is part of the problem.
I also suspect that humanity is too noisy to be a good source of training for these engines because we're all over the map on just about everything. If the Ai is generalizing from that data, enough people with weird preferences may distort the recommendations.
Thanks for chiming in. You have given me lots to think about. Will be back with a more considered reply later
👍🏼
There was a reason why it took me so long to get back to your comments; because I agree with most of them completely. Any AI based on language learning models will have the same structural difficulties so their output also should be checked. There are some people experimenting with a different type of training based on children but I think it will get more human dependent Increasingly some neurosymbolic approaches are cropping up at the edges.
I agree with you that it is too late and for the reasons you cite. Months ago I was chatting with Neela about doom and she thought we agreed. Her version was some bad tough years. Mine was real doom civilization collapse.
I here what you say about the recommendations engines and that perhaps human patterns are difficult to interpret. But in the scientific area they recognize very complicated patterns. I think the commerce motive tied with nostalgia humans overvaluing convenience is why they are successful
I did get a couple of comments that I cannot understand about decision fatigue. I hear people saying there are more decisions to make (which I don’t believe) and live is more stressful. I agree on the work / vocation issue it is. I think the real issue is right in front of them: an extra 3 plus hours dedicated each and every day to scrolling, with a million data points hitting your brain is tiring that what we used to do, which balanced complete downtime with the sustained focus of an activity like reading, listening to music, et al. These aren’t fatiguing activities; scrolling is
I do think we're beyond pendulum swings and see little hope for the immediate future. I think there's a strong element of boiling frog here. Things have changed slowly enough — and modern life has so many kinds of distracting soma — that people just don't realize that not only has the train left the station, but it's a damn way down the line by now.
I wonder if the science applications of Ai are more successful because those applications involve "dumb" matter acting according to physics. With human culture there is the extra component of intelligence (such as it is). I've long thought it ironic that what we call the "soft" sciences are actually incredibly challenging because of the preponderance of variables and human will. Many aspects of physics are extremely difficult, but we make progress in them whereas many aspects of humanity remain stubbornly opaque to us.
But, yeah, I think commercial motives are at least as big of a problem with recommendation engines as the difficultly of parsing human behavior.
Regarding "decision fatigue", I'm surprised you don't believe modern life is faced with vastly more decisions required of us than in the past. TV in the USA used to be three channels broadcasting a fixed schedule. Then there were local channels. And UHF channels. Then cable and now streaming. Movie choices have also expanded from a few theaters to megaplexes to VHS and DVDs and now to streaming platforms. We pick when and what to watch with few restrictions. We have access to an incredible amount of reading material, much of it free. Books, magazines, websites. Not to mention all the social media one can get into. Or consider all the shopping options we have today (e.g. an entire aisle devoted to breakfast cereal). We're bombarded constantly with people trying to sell us their product or get our attention. Myriad causes demand our support (and donations). Even just here on Substack, there are many blogs to choose from.
I quite agree we exhaust ourselves with mindlessness, but I think it's also true that the choices available to us now are dizzying.
I agree wholeheartedly with your frog boiling metaphor.
I think of the concept of complexity as applied to certain types of multivariate situations in terns of causality. So human behavior is one set, but the underlying nature and causes of the many diseases we know as cancer is another. As is climate despite the media’s gross oversimplification around “carbon”
I hear what you say about decisions (vs choice). For us, we have seen the increase in the numbers for some things which should increase our decision fatigue. But it shouldn’t for those borne into it; they should be able to develop strategies to deal with it. Like I have: for most consumer products, say toothpaste and mouthwash, there is no qualitative difference other than do I like the taste. So Find one and keep buying it. But in some cases there is less choice. Like restaurant meals. Three examples. Every pub in North America basically has the same 20 items on the menu, and when one makes a change (to ranch dressing they all follow). There are few creative, exciting options. Second, in every main shopping street in Europe or North America you have the same 25 multinational branded stores. This makes decision making much simpler and they are identikit websites for your digital shopping pleasure. No more unique shops offering some really different product options. Third example, when I was in my prime working years living in Vancouver you could get Ukrainian, Hungarian, Czech, Portuguese, Scandinavian, and 6 types of Chinese. No longer. The restaurant choices are greatly reduced.
If older people have decision fatigue maybe I understand it. But I think the cause in younger people is something deeper and more pernicious.
I don't at all doubt the influence of "something deeper and more pernicious". I'm not sure what to think about younger people who're born into this culture. Definitely there are strategies to deal with it, but even needing those strategies is a resource drain. A higher-energy way of life. Perhaps it's simpler to say that modern life is more exhausting. I just read someone point out how we combine activities now: listen to music while cooking, listen to books while driving, or talking on the cellphone while walking. Too many distractions to ever slow down and just be. So many things to pack into a day.
Maybe it has to do with where we live, but my experience of places to eat is variety. Even "burger joints" offer different sides, kinds of burgers, beers, and vibes. There still are unique shops, though a lot have gone out of business due to Big Box stores. FWIW, I suspect that as 3D printing becomes more of an on-demand shopping option, we may see more variety in products.
Oddly, I'm not sure the lack of variety in culture (which I agree there is truth to — all the chain stores and franchises and movie and book sequels) detracts from there also being so many choices. Even when some channels lack variety, there are still a lot of channels.
The little door was never a problem when it led to fantasy conjured up by books or other art. Somehow that only enhances reality or makes it more bearable. But with technology, we tend to confuse the two states. Why is that?
You are right. But the metaphorical door in technology is profoundly different. I really admit to not understanding why people want to overvalue techie experiences . Just saw another study that found that 45% of under 25s were online continuously
I would say that yes, the majority of us have decision fatigue and want to be told what to do. If that weren't the case, never mind tech, COACHES, wouldn't have a business model!
I realise that when most of us go to a coach they try to guide us to realise the idea ourselves, but a lot of us contract those sorts of services to be told what to do. Damon Mitchell wrote about this recently but I can't tag in comments so I'll send this to him 😊
All this being said, I reckon we all assume that these tech recs come from human input, at some point... We are ignorant of how much is actually machine learning and AI, perhaps deliberately so.
I am interested in the issue of decision fatigue. Which decisions and why the fatigue? If it is around career / vocation / job nexus, I can understand as this part of our life is different and more difficult than in past generations.
All of AI is some statistically moderated version of human information
Personally, it's at work having to say yes and no and maybe to everything my team asks me, also to take in suggestions from other teams and decide which to listen to. At home, then, just deciding what to eat becomes a chore. People who wear and eat the same every day male perfect sense to me, as a result. They make a handful fewer decisions than I do every day 🫡
Thanks for sharing about this. It has been 25 years since I was in the daily operational firing line where decisions were required constantly. I menu planned on Saturday then the decisions were made at a down time. But i love cooking so it helped me get out of my head when I came home.
I could not menu plan to save my life at this point 😆 one day I would love to have the mental bandwidth as well as skill to do that. I rotate very simple meals in my home, partly she to lack of creativity but mostly lack of ability 🍳 any suggestions for a Level 1 Menu Planner?
I laughed at your comment as my wife always says I should write a cookbook for exactly that: healthy, tasty meals that are ready in 10 / 15 / 20 minutes because 2 of our 3 kids and their partners are in panic every dinnertimr. I have a lot of them. Maybe I should.
Thanks for your input. Thus article generated lots of different perspectives. I want to think a out what you said and will het back to you again as I wanted to ask some questions. But now my granddaughter’s demanding fun on Canada Day
Thanks, Nadine. (I'm also unable to tag you.) I'll not share my essay here, lest I seem to be attempting a hijack of the audience. Smart people can find their way to my Substack if they feel so motivated.
As I read this essay, I thought, "Yeah. These engines of choice have never worked that great."
It's ironic that in this world, where GenAI seems to be speeding us along to every valuable conclusion, we still can't help folks figure out which content is available that they will actually want to consume. Just yesterday I started a ChatGPT convo to aggregate my preferences and three movie/series review sites to see if I could hack this more locally. The jury remains out, and I remain hopeful and skeptical.
But I'm not sure I want my GenAI creating more efficiency. Every time I chase efficiency, I have to trade out the wisdom of experience to get it. This is why I don't own a commuter mug or a car to put it in. It's why I don't walk and text, and generally downthrottle every intuition that multitasking is somehow better. It's why I live in Latin America, where "mañana" doesn't mean tomorrow. It means "Not today, Gringo."
My hope is that the backlash to all of this is a kind of slowing down. I'm sure it won't be, but I still hope for that.
Thanks for the comments. I read your latest article. I found your crowfunding metaphor for Chatty GPT refreshingly naive. It may have ingested tomes or Epictetus and the Dalai Lama but also tens of thousands of uninsightful nobodies. It’s massive statistical engine generates primarily mediocre and at times average outputs across a incredibly broad spectrum. I have found with some experiments on subjects that I’m an expert in that AI is an unreliable Encyclopedia with absolutely no real understanding of priorities or preferences. And AI, as currently constituted as LLM‘s are architecturally and conceptually unable to get any better. It has no way of moving toward truth, so despite the underlying technology being known as machine learning, it’s really just doing statistical association of words. The other thing that I find that it does which you mentioned in your peace is that actually disrupt human connection. To me that is a profound negative implication of this technology. However, the conflation of so many other technologies are mean that it’s just one more disruption on the dehumanisation of technology.
Honored that you read my essay. Thank you, David. I'll take "refreshingly naive" as a compliment. At fifty, "bitter and curmudgeonly" are closing in faster than I'd like. Naive, hold me tight! 😉
You're absolutely right, but we all know fast food is not very nutritious, and yet here we are. Even if current models are aggregating bad intel, it's cheap, and it's convenient. I don't think anyone cares or knows enough to care, whatever you and I believe they should do.
Thanks for your reply. Curmudgeonly is a cheap ad hominem attack especially when applied to oneself. I think bunches of individuals using AI for free are incidental to where the technology goes, from an economic standpoint. If it doesn’t work successfully for business organizations (and the real data shows an excess of problems with few successes )the size of the investment is unsustainable. If I could sell OpenAI short I would.
You mention special purpose AI doing real good, but it's hidden behind the hype of general purpose LLMs. This feels like a microcosm of the larger issue - appropriate, measured technological solutions exist, but they can't compete for attention against the flashy alternatives. It's not just that we've become technophiles, but also we've become addicted to technological spectacle. Maybe the solution is better media literacy and a cultural change. Access to information does not equal understanding.
Happy Monday evening, David.
Thanks for your input on these issues. Not sure what will happen going forward. There is a strange fascination with tech spectacle which is the cousin of celebrity worship. We have the growing problem of tinpot ChatGPT experts who do one prompt and bingo!
Yea it will end badly but sometimes I think we have to go down this road David. Burn to learn - something like that. The problem is we NEVER learn.
Sad. It is the never learn that DRIVES ME CRAZY. Great song from the 80s. We never talk to others. We never look back at history. We continually reinvent ourselves like cigarette smokers from the 1950s when we knew smoking killed……..each individual : but not me