I am reminded of Veritasium: What Everyone Gets Wrong About AI and Learning (original video https://www.youtube.com/watch?v=0xS68sl2D70) from a year ago.
"The world is full of heavy things, and yet most of us aren't ripped."
AI is an opportunity. On the one hand, it can be used to let our minds and social lives atrophy. On the other hand, it is an opportunity to help our minds grow. Most people will make the lazy choice. But you can choose to do otherwise.
Take, for example, speeches. I do not let AI write my speeches. But my speeches are better for having been critiqued by AI. But the result is still my speech. My thoughts, my ideas, my words, and my meaning. Just improved with rounds of feedback about where it fell flat, where I was likely to lose people, and so on. Feedback that I had to fix.
So do not let AI write your speeches. But do use it to push yourself harder.
I had this moment when we designed shirts for the marathon we ran as a group. Instead of Brainstorming something funny, we just prompted ChatGPT and chose one of the results.
I felt lost immediately. All the creativity, the humanity, the endless hours of putting soul into something. Gone
For one hour or so I had some kind of existential crisis. Just because of a funny slogan on a shirt. And sometimes I still feel empty on new projects. You can produce so much things so fast, but if it should be something original - it is hard to get it generated by AI while still feeling that it is something that you came up with
Ever since I started experimenting with AI coding, I've totally lost that feeling of accomplishment. For projects I actually developed by typing in the code, it feels like I actually did something--like here's something I built and am responsible for bringing into the world. When I finish an AI-built project, I feel...nothing. Just that empty: "Code now exists where it didn't exist before, but I didn't really do anything." Without any sense of ownership or attachment whatsoever. If someone DMCA'ed one of my GitHub projects and made me take it down, I'd be pissed. But if someone DMCA'ed an AI-coded thinggy, I'd probably delete the repo and never think about it again in my life.
This makes me think about that "Dad, how do I?" YouTube channel that made headlines a few years back. People seem to be fine with such a thing existing, they don't seem to be lamenting that people might go to that channel instead of asking their own fathers.
Like, apparently Mr. Smucker has a friend who's into fly fishing, and the time to talk to that person. Great! Good for him! If I do not have a friend who's into fly fishing, or if I need an answer quickly, am I...just out of luck?
I understand the impulse behind posts like this, and it's important to remember to maintain human connections. (Arguably, once we learn how to do this because we think it's a good in its own right and not because we have to, we'll be better off.) But I just don't like being emotionally browbeaten like this because I have a question that I need an answer for that I don't have the time, money, or access to go get in a different way.
I really don't understand taking the author's silly hyperspecific examples of unique humans in his life as berating the reader for not knowing exactly those same people. I read it as "remember all the unique people you know and try reaching out to them instead of going to AI or the internet."
A lot of people don't have that many friends. I forget the average but it is in fact absurdly low, at least for Americans. There are a lot of reasons for this (e.g. erosion and disappearance of "third place" spaces, rise of social media, etc.) but the circumstances have essentially been ripe for something like AI to come in and fill the gap, and it is.
>If I do not have a friend who's into fly fishing, or if I need an answer quickly, am I...just out of luck?
I really don't understand the need to torture alternate meanings out of the writing of people we don't agree with. Nothing in the author's writing even comes close to implying what you're suggesting here.
There's an undercurrent in a lot of writings like this that don't seem to grasp that LLMs enable access to a ton of knowledge that was otherwise out of reach for a ton of people.
I'll give an example. I just traveled to Serbia, and I went on a run through a park in New Belgrade, where I saw a monument written in Cyrillic. I snapped a pic of it and uploaded it to Claude; it translated and gave me some context.
I thought this was amazing!
But I'm sure someone could point out that I took a mental shortcut, that I made myself dumber by not grasping Serbian and Cyrillic to have a go at translating myself. Or they could say that I lost the human connection that would have come by finding a resident who spoke English and asking about what that meant.
In a sense, this are plausible critiques. But the reality is that I was on a run, and I almost certainly never would have done those things if Claude didn't exist. I didn't become lazier or lose the imperfections of human connections, the whole thing was a net add for me.
And so, in that light - it's okay to use a recipe book, or ask an LLM about fly fishing, or do some web searches to get some advice about how to write a wedding toast.
If that's missing the point somehow, so be it. Perhaps you could enlighten me (and thus cultivate a human connection)!
I found it rather on point to be honest.
> This makes me think about that "Dad, how do I?" YouTube channel that made headlines a few years back. People seem to be fine with such a thing existing, they don't seem to be lamenting that people might go to that channel instead of asking their own fathers.
Mot everyone has a father to ask. His own family were abandoned by their father when he was 14 and his sister was 9. People die. Some people have abusive or neglectful parents.
Not every dad is good at everything.
> This makes me think about that "Dad, how do I?" YouTube channel that made headlines a few years back. People seem to be fine with such a thing existing, they don't seem to be lamenting that people might go to that channel instead of asking their own fathers.
Didn't that guy start his channel because he didn't have a father growing up? Seems like important context.
Right, that's my point exactly! Sorry I didn't mention it.
It's a channel that increases access to knowledge for those who wouldn't otherwise have it, but disrupts a status quo in a way that some might find harmful. But in that case people seemed to pretty universally recognize that the pros outweighed the cons.
> disrupts a status quo in a way that some might find harmful
I love a good strawman argument myself, but this is just madness. Who the heck finds substitute "dad advice" harmful?
Analogies are almost always a distraction.
A YouTube channel about stuff your dad might know does not have the same potential for negative impact on human interaction as genAI. And the author never even claims "the cons outweigh the pros". Maybe they feel that way, but the dangers they advise against are absolutely real and do not require a broad stance like "everybody who ever uses AI should feel bad" in order to recognize those dangers. I use AI every single day, yet I do not feel the least bit browbeaten and my heart bleeds in agreement with this blog post.
Well maybe if he didn't spend his time on YouTube he might have had a father.
You underestimate how easy it is to get someone who's into fly fishing to talk about fly fishing. You don't need to have known them for more than thirty seconds.
Even NYC has a fishing meetup group with over 1000 members.
It's harder to get them to stop.
I love when I get someone to talk about something they clearly love, and they're giddy with joy and struggling to contain themselves. It's one of the finer pleasures of talking to strangers and not machines.
Good point. Fixating of the fly-fishing example is silly to begin with but yeah- if you don't know a guy, it's certainly an opportunity to meet one.
> If I do not have a friend who's into fly fishing, or if I need an answer quickly, am I...just out of luck?
I know, right? The author clearly wants you to starve to death for the lack of a friend to teach you to fish
Eh, the poem doesn't suggest technology isn't ever useful. It's highlighting that the inefficiency of human relationship is a feature, not a bug.
You might not have a friend who is into fly-fishing, but surely you know somebody into SOMETHING you could ask about. Maybe that's less efficient, maybe it's less direct. But our whole reason for existing, all of the stuff that gives life meaning- it requires each other, and technology is getting dangerously close to replacing relationships altogether.
I don't think this is meant to guilt you for using tech, but it is totally a wake up call to remembering WHY we fly fish and go to weddings and write memoirs and so on.
Way to miss the point, there.
>they don't seem to be lamenting that people might go to that channel instead of asking their own fathers.
Much of the anti-AI sentiment has this sort of false dichotomy as its foundation. An imagination that the alternative to AI is the purest form of manual labour in some sort of idealized, bucolic form, filled with heartfelt, purposeful, sincere human connection.
So every time I'm thinking about what to make with the ingredients I have, I should text someone who cooks (I cook, so this is a hypothetical)? What a ridiculous canard, and absolutely no one would appreciate that. I can enjoy human contact without inventing ridiculous justifications.
Further, to quote from Unlearning Economics, everything already was AI [https://www.youtube.com/watch?v=Km2bn0HvUwg], at least in the demonized way that people use that phrase.
Wedding speeches? Overwhelmingly cliche bullshit, and if you've been to a number of weddings it starts to get incredible how blatant this is. The whole manner of "genres" of music, art, and so on, is everyone copying each other and mimicking styles.
Even the recurring "I can spot AI websites!" nonsense, as if everyone wasn't already copy/pasting the trend du jour.
Even programming, this site is stuffed with "I lament the loss of the craft" pearl clutching articles daily, yet most of you are terrible programmers. I mean this as nicely as I can. It's astonishing seeing the actual state of the industry and hearing people imagining the world's most skillful, conscientious, thoughtful developer as the only alternative to AI assistance. It's rather amazing.
And long before AI people were largely just duct-taping together whatever libraries they found mentioned in a StackOverflow post.
Is it possible to hand craft better creations? Absolutely. Was that the norm pre-LLM? LOL, not even remotely. People were churning out enormous volumes of garbage, in every field.
AI isn't the reason people aren't making "human connections", and the foundation of the article is perverse.
>Much of the anti-AI sentiment has this sort of false dichotomy as its foundation. An imagination that the alternative to AI is the purest form of manual labour in some sort of idealized, bucolic form.
This is backwards. This false dichotomy is what irrational reactions against anti-AI sentiment use, not the anti-AI sentiment itself. It is exactly the false dichotomy the parent you are replying to is using.
Beautiful piece.
I sometimes feel like technologists actually desire to remove the humanity from the world because it's messy and they don't understand it and therefore they fear it.
I feel the same too. And I believe there is much more complexity in the question "will this be good for society overall" than technologists can apprehend. For example even though I recognize some benefits to social media, I'd have a hard time arguing that on a societal level it's not a huge net negative. Overall, people are more divided, more angry, depressed, egotistical because of social media and the attention economy. And ultimately, as one of my previous boss would say "it's all about people".
> they don't understand it and therefore they fear it
I feel this whenever discussion of consciousness comes up. Even though consciousness is not well understood at all (e.g. no scientific progress whatsoever on the "hard problem"), some people would rather say "it's just molecules and we don't have free will, we don't really exist, it's all an illusion, science will reduce it eventually, etc. etc." It baffles me that some people would rather contradict their very experience and declare that they don't exist! Rather than admit there's something that may be impossible to understand.
> "Technology [is] the knack of so arranging the world that we don't have to experience it."
-Max Frisch
The whole idea of trans-humanism, so beloved by VCs and the AI cult, seems borderline psychopathic to me.
The poem is absolutely on point. Nobody wants to consume AI content, especially on the parts that should be all-human.
At the same time the poem is published on Substack, instead of a hand-crafted custom blog.
There are 1) the tools that let us surface the human, then there is 2) the human, and then there comes 3) the factory generated business (someone doesn’t care but has to do it) content pretending to be human to sell stuff to humans. The human 2) is drowned out by the “had to do it” 3) while there is a small corner of some of us who are making 1) tools to surface and reward more 2).
> At the same time the poem is published on Substack, instead of a hand-crafted custom blog.
Look. I am a massive fan of the janky old manually created website. <marquee> will never die and it is hilarious that browsers will have to retain the feature for years to come.
But "the blog was generated by a machine" isn't the problem with Substack. "Machine Generated" blog sites have been around ever since blogs went big. Blogspot and Wordpress were practically a duopoly in the peak days of blogging. The problem with Substack is two (really, only the latter):
1) It's gotten the Post-Zuckerberg "everything must follow our company letterhead" disease. That's not a substack exclusive problem and designers need to be bullied harder for it.
2) It's the nazi bar where all the nazi blogs are. This one is the actual reason you should not be using substack.
The "If it's on substack, it's not a real blog, it's just sparkling page bloat" take is a little strained for me.
The content is great. The tool gives writers a low-friction mechanism to charge for premium content, and works on most people's devices. I would rather have read this on substack, than the author get frustrated at having to learn how to publish pages by hand and give up.
Maybe we just need a better alternative to substack if that's the problem.
Being great writer and capable of self-hosting your blog is a pretty unusual combination once you venture outside of the realm of tech.
It wasn't 20 years ago, and it shouldn't be today, but somehow we've made it harder. I suppose some think AI will "fix" it but I tend to think it'll just make it worse.
> instead of a hand-crafted custom blog
I think this kind of elitism also misses the point.
So what's the kind of elitism that gets the point?
This does not just apply to AI. Uber, AirBNB, Facebook, etc. all basically serve as paid surrogates for what once was done by community.
Sometimes it feels like all digital technology is simply an enterprise to replace human to human contact.
> Sometimes it feels like all digital technology is simply an enterprise to replace human to human contact.
Hasn't it always been the case that technology reduces the contact with other people? Now with cars we don't need to sit next to others on trains, we don't need to ask pedestrians for directions thanks to GPS etc.
Facebook, sure, but Uber and AirBNB? I don't see how Uber has displaced some community function. AirBNB is arguably destructive to communities, but again how was community fulfilling the need it attempts to address?
Before Uber it was totally normal to ask someone, even an acquaintance, for a ride to the airport.
with classified ads? or calling the local tourist office? Like people didnt rent a house for their holiday before airbnb
Most people did not rent houses on trips before airbnb
This was great. I think about this a lot and have for years now.
When LLMs first showed up I thought “but doesn’t this take away a little bit of what my life is? Don’t I like programming and solving the problems and learning the unexpected things and so on?”
Now I use them extensively, daily, millions of tokens per day, and I still ask that question.
I don’t use them for recipes or toasts or camping trips. I use them for brute-forcing boring stuff. Like, hey we’re making this thing faster. Let’s measure all this stuff, and you come up with whatever I’ve missed to include in benchmarks. Make a benchmark harness for each approach we’ll try. Create tests to ensure none of the changes alter behaviour or outputs of the system. Make it pipe results into this database with this schema. Let’s try these approaches. Which other approaches could work? Keep slamming these benchmarks until statistically significant results appear.
The thing we’re speeding up is usually a single query in the armpit of an application that in prior years I never would have been able to address. But now I can. By doing this I can improve the user experience and scale back our resources and other stuff we like.
Am I missing out? I don’t know. I program less. I get a lot more done. My employer is very happy. My team expresses appreciating my work more than ever. It’s a stark contrast, actually. It feels weird.
I’m still not sure what the answer is. I do miss tinkering. Yet I suppose the point was never me tinkering. It was me having a job to perform for a specific purpose for my employers.
Did it take away a bit of what my life is, or did it change it? I’m still using my brain. I’m still thinking through problems. I’m still finding bugs and mentally tracing them to understand how to work through it with Claude. But the actual moving of bits? I don’t do it anywhere near as much as I used to.
I’m still very conflicted about it.
I’m so disturbed when I see friends and family using AI for ‘real’ stuff. Recipes, images, writing, etc.
Is programming ‘real stuff’ too, though?
To be honest, when working on personal projects with AI I feel like I've replaced some of the joy of tinkering with code with the joy of tinkering with models. They require different work, writing prompts, setting up guardrails, harnesses etc, and that is also pretty fun for me!
I just use it as a "mentor". A captive demon that has to answer my questions, no matter how trivial. Writing the code is the fun part for me, searching for answers to questions can be fun but I'd rather just ask the AI. I even ask them to give me longer answers so I have more context, even with languages I've worked with for decades.
Did you use AI to write this ? Feels like you did.
If AI is not that special, just a tool, then treat it as such.
If AI is special, unlike any other tool, why aren't you using it that much?
I personally don't think it's anything special, and if I knew I'll die soon and were planning my last trip with my child, I'd use AI, just like I'd use a credit card, or my phone.
It allows me to spend more time with other people, getting boring tasks done much quicker.
Hypocrite didn't even use AI to write this lovely poem.
OK, but you could write the same thing as "Please read books". Many times I have learned things from reading I could have learned, e.g., from a crotchety old neighbor in return for interacting with him.
Soulfully! AI just a tool. Person constantly uses the tool instead of itself is felt as a robot.
Beautifully expressed. Using AI to remove even more opportunities for human contact is a tragedy.
We're optimizing the soul out of being human.
I don't think it began with AI. We repeatedly catch the car we're very deeply programmed to chase. We want to minimize discomfort, risk, suffering, adversity. We want to maximize safety and comfort. We want all of our kids to make it to adulthood. We want to disinfect the planet of all diseases. We want our bodies to survive a career. We want our families to survive every winter. Those goals are all completely sensible.
But parents, for example, have been here before and recognize that optimizing these sensible goals have a consequence of missing the richness in the journies we no-longer need to take. So have those who have grappled with social media addiction or the withering effect of sedentary careers, or even the little things like waiting at the radio for your favourite song, your finger hovering eagerly over the record button of your cassette player.
I think this is going to be the supreme challenge. We're wired to seek the destination of comfort, but we lose the journey to reach it. It was easier when we had no choice. But we're doing a great job optimizing the soul out of being human.
I know a few guys here who were doing sysadmin, devops, frontend jobs for a few years in India and now they are driving a taxi in India.
AI took their job. There have been mass layoffs by foreign companies in India; fewer outsourcing contracts are flowing to India.
As a result, many service companies are moving to product businesses.
I've been pondering the question: "What does it mean to live well with AI?"
We are certainly scrambling for productivity with "token maxxing" and scrambling for entertainment with AI companions, but I haven't seen many thoughtful takes on how AI might look in a life well-lived.
Even though I spent the majority of my work day with computers, my fascination with them starts and stops on understanding how they work. Aside from that, they’re only utilitarian. What I really like to do is grab a nice book, put on some nice album to listen too, and enjoy a quiet night with my SO. If not for the fact that it’s easier to get books and music in digital format where I live, I’d spend even less time with computers.
Reminds me of that silly Adam Sandler movie Click (2006).
In that movie only the protagonist had the magic remote to fast-forward through existence. It was a tragedy of self-destruction.
But what if everyone gets the remote at roughly the same time?
This movie hit harder than my highest expectations from an Adam Sandler movie.
Be sure to use a mobile phone when making your next, I don’t know, meal plan, for example. Definitely do not come in person to your friend who loves to cook and ask her for her favorite recipes or tips or ways to save time making meals
You've missed the point. You're describing replacing a method of communication, the poem describes replacing a relationship.
This is really beautiful and tragic at the same time. Very well written.
I thought I'm jaded, and a bit of a poet myself, and already sufficiently "upset" by several things, but this still made me so profoundly sad, and at the same time incredibly proud of the author and hopeful of being human. And they don't cancel each other out. It's a very strong, odd mixture. This is with me now, and I hope it'll linger.
I don't have anything intelligent to say really. This poem made me go "Fuck yeah, poetry! Humans!!", and I'm grateful to the author, the submitter, and the people who upvoted it, so that I ended up reading it.
This is beautiful.
Or just use AI when it makes sense, and call your friends too. Why do we have to over-dramatize everything?
I don't see anything over-dramatic. He's writing about a real problem affecting real people, and he's not exaggerating. Just because you believe you are balancing things properly doesn't mean everybody should just shut up about it.
The problem is that we have incentivized efficiency over authenticity even in our inter-personal relationships. It's a systemic issue. It makes it very hard for most of us to resist the sirens of "let me just rephrase this important message so that it sounds more elegant/well-written/relevant/...". In the current cultural and societal context you need to swim against the current to _not_ be using AI for everything. So I don't think this is over-dramatization. Overall, on a societal level, we truly are moving in a direction where we are robbing ourselves of real, authentic moments by using AI because it's "convenient/efficient/easy/etc...". Even at work.
Many people don't know "when it makes sense". This highlights when it does not make sense.
I think its fascinating how many people in tech think there's a clearly defined and agreed upon "right way" of using this technology that everyone knows and abides by. Paul Graham, for example: https://x.com/paulg/status/2058871512451412457
It's like we memory holed the last 20 years of social media that was supposed to be all upside; just democratic, global connectionism, empowerment, etc. I have too much exposure to people using AI in various, even sometimes subtle "wrong ways" to really agree.
"Ten scenarios that I invented in which AI is making my life miserable."
The same predictable comment comes up whenever there is a piece that isn't sanitized, blunted technical documentation. Why write long form literary pieces that take effort to digest when you can get a cliffs note. Why write poetry when you can write a tweet. Why have anything resembling anything with humanity when there are summaries and machine written slop.
This sort of comment plays exactly into the thrust of the piece.
Or you could use AI to explain to you how you missed the point.
I'd have thought people that are technologists at heart would have understood the benefit of the next Industrial Revolution but all anyone wants to do is whine about it.
even if this stuff is the "next Industrial Revolution", the Industrial Revolution was famously Not Good for many, many people
I see this false equivalency argument everywhere. Just because one revolution had one effect does not mean they'll all be the same.
Nor does it mean people negatively affected at the time were wrong to fight for the quality of their family's lives. Anybody would do that. Also, the unfolding of inevitable changes can be managed by governments to reduce harm (they just usually don't because that would mean a slower increase in profit, directly or indirectly).
I agree with this sentiment.
I've seen other parents create AI videos of their toddlers being visited at night by Santa. I've seen parents happily throw their children into AI video generators to entertain them.
People are using AI recklessly. I can't imagine stealing the gift of a child's imagination away from them and instead, replacing it with these hollow representations of reality. It disgusts me.
I use AI all the time for coding, but I've drawn a hard line at the point of intermediation with others.
Really beautiful piece.
I sometimes wonder if these same people pushing AI onto devs would ask the same of their lawyers and accountants.
If someone hasn't gotten the memo yet, writing code got that serious at least a decade ago when web ate the world and chrome had won the web. Probably even earlier for certain industries like financial institutions.
This isn't just about "human imperfections" or something else sentimental. It's the fact that quality really does matter in a huge number of situations and the consequences are not forgiving in the slightest.
Misses the mark IMO. You can already do all of these things. Just do them. As long as I get to fire half of my employees and you hit the token quota it’s all good.
Wow, so powerful, I could barely type this comment with tears in my eyes.
OP should consider a side career in poetry.
Beautiful.
Thank you!
Every interaction I've had with AI has been negative. It's just not very good.
I really dislike the condescending tone of the person who thinks they discovered the secret of happiness, but instead of distilling their wisdom for joy chooses to shame others in a passive aggressive poem.
Sure, buddy, you know how to live a meaningful life, then why are you trolling the internet?
It's a real shame this somewhat interesting tech is entirely under the control of the most insane, inhuman, sociopathic monsters produced by our modern society. There's lots of genuinely cool, interesting uses for this tech, but instead of exploring those uses, the monsters have used it to drive a wrecking ball crane into the middle of our society and then call us morons for not saying "thank you sir, may I have another round of beatings, please?" as they tear down everything our society worked to build over hundreds of generations and plunder the copper pipes from the wreckage. Whatever uses the tech may possibly have had, they're all tainted with the stench of the 5-day-old corpse the tech bros keep shoving into our faces and telling us is the only food we're allowed to eat, and I want nothing to do with any of this AI/LLM crap.
I asked Claude what he thinks about this blog post and was surprised by the level of self awareness (you cant call it like that but I dont have better word)
Meh, here's a haiku from gemini
> write a haiku for stop using AI for human things and use it for automating the boring stuff
Let humans create,
Leave the soul to living minds,
Let code do the chores.I really love it when people put spirit into a piece of writing that, thanks to an algorithm (that's another name for AI, by the way) suggests it to me on HN.
I am pleased that I can share musical discoveries with friends that were recommended by an AI, or make them laugh with some absurd image that fell out of Dall-E.
I am happy that, with the help of an AI, i can make a news reader that is full of bright patterns, instead of dark ones, that i can share with my friends so that their standard of life is ever-so-slightly better.
Reducing the commentary to "tool bad" is lazy, even when beautifully phrased
The author’s point was more nuanced than ‘tool bad’.
> thanks to an algorithm (that's another name for AI, by the way) suggests it to me on HN.
It's a pretty big stretch to liken a ranking algorithm based entirely on direct, intentional human inputs to what most people understand to be "AI".