I'm glad I finished my Ph.D. before any of this stuff hit the market. It sounds like it would have been a substantially different experience, and I suspect I might not have finished it with as solid a grasp on my thesis subject as I have.
Her comment about the Mathematica notebook really hits home. As a scientist, not a software engineer, who uses code as a tool to speed up more critical parts of my work, I was thrilled to have the chance to learn programming skills on the job. I could justify it because the time saved by the fruits of it outran the time I spent learning it. I can't do that anymore. I'm not building software, just scripting, so even the cheapest models today can give me suitable code quicker than I could write it myself. So I can no longer justify learning new programming skills, at least on work time. I'm glad I had a few years to pick up the basics before these tools became available.
It’s the same at work. You don’t ask your colleagues for help anymore, you ask the agent. You can clone all the necessary repos, link to confluence/figma/drive and then ask.
It sucks. I don’t care about being efficient (that’s something the executives care about). I care about how I live my life, and at work that means learning from others (and offer the little knowledge I have to others)… if it takes 2 engineers 2h to solve a bug (vs 20 min for 1 engineer and an agent) then thats great for me. I get the chance to exercise my brain, solve a bug, learn from other engineer and have a bond with that individual. If I solve the bug in 20min with an agent, well, that’s something my boss would love, but I personally would have gained nothing… and would have to jump to solve more bugs alone in the remaining 100 minutes (which feels like running a marathon)
My anecdote is that most of my coworkers across multiple jobs didn't like to explain things. I think most engineers don't like being interrupted for questions and I would have appreciated a way to get answers without bothering them.
At the same time, there's a type of person who loves to explain things. They like to give presentations and are extra helpful when engaged. I'm pretty sure you can still ask them questions.
Problem is that the type of person that leans towards collaboration and likes to explain things may be getting burned out right now or may be among the first to be replaced.
OTOH, it's great for the type of person like me, who likes answering questions but is rarely asking them. Like, I don't want to interrupt my colleagues for something I can figure out on my own in comparable amount of time/effort. AI is now scaling this to questions previously way too difficult to answer on my own.
Anecdata: I am that person, loves explaining things, love seeing my colleagues learn things and grow. Needless to say, I no longer hold a job in the industry.
I maybe liked to explain things, but never liked to ask other people, always liked to figure things by myself. Maybe it was the experience from school and university with a lot of teachers that didn't know what they were talking about, which made me not trust other people.
I'm the same as you. I hate asking other people, unless it's for a reference or material to learn. I like to grind to understand myself, and then answer questions and explain things via presentations or a drawing board. I never wonder why I'm like that, tho
That said, hardly anyone asks for explanations now. In my team people default to a chatbot for work questions almost exclusively, and in private life for hobbies I share with other people, they increasingly go for chatbots too.
Turns out, not everyone seems to want a human connection.
Oh, we do want a human connection. But we also want good information, and want to actually know this (i.e. hobby is something to do, not merely a pretext to socialize). The overlap between the two is low.
Efficiency in the short-term also leads to community degradation in the long term. And by efficiency I mean efficiency past the point of diminishing returns. Compiling code by hand would suck, but not writing code at all also sucks.
There is no formula to determine the right level of it, but it's one of those things that's obvious when you see it... like not eating.
Neither do I. But what I suspect is that the unpleasantness of too much forced social interaction, which dominated discussions in times past, was also associated with the begrudgingly-acknowledged good it did, even for us introverts. Hey, I'm like that too. But AI has taken that away for many people, who are now realizing that occasionally eating those dark greens actually did some good and now they're gone, life is actually worse.
This stereotype of "hackers" feels very rooted in the late-90s/early-2000s. Ever since mid-2000s the prominent "hackers" have been collaborating in the open and chasing VC funding...
Not every hacker fits the stereotype of "guy typing fast in a basement filled with screens and terminals".
People are people, all different, some hackers enjoy the solitude of doing their project by themselves with no human interaction whatsoever, others enjoy and thrive with a collaborative setting, spitting ideas, showing code, and getting feedback to iterate on and improve them.
I've belonged to the latter camp most of my life even though I acted like the first when I was learning how to program (and on my first job); I'm not a stereotypical extrovert but love nerding about stuff together with someone: coding an idea, making music, working on wood, for me those are all more enjoyable with another human sharing the same passion.
Hackers are not anti-social, we're just a subculture. Hacking itself is ultimately a community activity. You want others to appreciate this cool thing you've made, or share tips about it, or just in general provide people connection. People buy plane tickets and conference tickets, pick their university/major, accept materially non-optimal job offers, just to seek out other members of this culture.
And for programmers that aren't hackers, they just appreciate the practical upsides of having mature humans that can take accountability for decisions that affect your work as coworkers.
> Hacking itself is ultimately a community activity. You want others to appreciate this cool thing you've made, or share tips about it, or just in general provide people connection.
hacking something alone is fun, but maintaining something that’s being hacked by someone else without coordination has always been a shitty experience.
Even those that have less bandwidth for human interaction value being in the loop.
Presumably the people who loved interacting with humans were too busy interacting to show up on forums. And now that interactions are being curtailed they do end up here.
I've discovered that I either want to have no interaction with people, or I want to have real human interactions. I don't want the shitty half measure where my manager sends me whatever Claude sent them, or gives me instructions like I was a prompt ("do X now").
I feel there are workplaces and career choices where those people go who want social contact and fulfil personal utopian ambitions, not necessarily having profit in mind, and jobs where you go to make money efficiently. And these days, the divide between those has gotten very large to the point many jobs have no personal contact or room for personal growth, interests or artistic skill anymore. It feels the sane middle ground for sane people who are looking for a sane balance has been or is being hollowed out.
And when that workday is done, everybody goes home by themselves to scroll some more.
Do you apply this logic to all efficiency gains or just ones that weren’t commonplace when you entered the industry?
For instance, package managers didn’t always exist. Would you prefer it if we got rid of them? Instead of 30 seconds running `uv add whatever` collaborating with your package manager, you could spend 30 days collaborating with other human beings to build the functionality yourself.
In capitalism companies do things what is good for the company.
You can try convincing them that doing the same job in 2hrs is better for the company but they may not listen.
Like in outsourcing when they say the Indian worker does the same job for a much smaller salary and they don't care about the quality of the code they produce only about hitting the milestones. If the code is bad then they hire some other cheap outsourced workers to fix it, etc.
The market does not reward good code, the market rewards low price and speed of delivery.
Time and time again, I see the same complaints. A new technology comes and it removes bottlenecks. The ideal person should be ecstatic with new capabilities being unlocked. But they are complaining. Why?
They don't trust their peers to use it properly and externalise the concerns to the technology itself.
> The problem is that the little chatbot in our pockets is making us more isolated and in so, it’s making us stray further from this humble goal: to learn
I don't understand the author here, ChatGPT (as admitted by herself) does help people learn. Far more than what was possible before. Her only complaint is that society is not able to leverage ChatGPT to further the spirit of physics, but that's on her on peers and not on the technology.
> In particular, it would be pretty sad to destroy one of the few things that makes our life meaningful, not the fancy prize or the fancy position, but simply having a community to share and discuss and learn.
Exactly. The problem is a weird one though, because there are at least independent forces that are working together here:
(1) The people that are inventing AI or those that are pro-AI are typically those that have been conditioned to enjoy the "information rush" and place personal superificial enjoyment over human interaction. Lots of nerds/geeks simply love the idea of having a playground devoid of the necessity to actual interact outside their microcosm and so they think AI is the best thing since sliced bread. They're often people who were taunted in school or at least partially ostracized early in life because of their intelligence and take emotional comfort in tech.
(2) The perverted incentives of academia reward publishing and in general the incentives of capitalism create an arms race to be as efficient as possible at the cost of goods like "community" that only accrue vague benefits often outside the capitalistic system. So people tend to use AI.
(3) Very advanced technology itself weakens and even destroys smaller communities whose members then seek out technology as a superificial salve to ease what they consciously or subconsciously feel as something missing.
AI was never great. Knowledge as an end in itself is not a good thing and we should seek knowledge to better ourselves in a myriad of ways. But today it's just sought out for material and capitalistic advancement, which prevents our using it to actually become wiser and enjoy the basics of life.
The only people who defend it are those who are already addicted to it, or those who don't want to experience the cognitive dissonance of the truth of its harmful nature.
I somewhat agree with the article on the narrative level.
Lately, my thoughts on our LLM future darkened a tad, too.
I find the parallel between LLMs and Sophons from the 3-Body Problem novels increasingly accurate. That is, specifically, in terms of their effect of stalling human progress. Sure, anybody can opt out of using LLMs (unlike with Sophons), so the restriction on progress seems kind of "soft" in this sense, but the chances of someone choosing to do so aren't too high, IMO. It is so much easier to avoid forcing yourself to learn or explore hard stuff and just ask the model instead.
While I tend to support this in a field such as SW Engineering (if only the Standards and Quality Control aspects advanced as fast as raw code generation...), I share the same fears as Tao and others regarding Mathematics (and not just Mathematics; Computer Science, which ultimately underlies SWE, is on the spotlight as well).
>It is so much easier to avoid forcing yourself to learn or explore hard stuff and just ask the model instead.
What exactly is hard stuff though? The concept, the implementation, the grokking? For a person with the right mindset now is easier to get to the Eureka moment, because it is fairly cheap to try all the stuff that doesn't work.
Reads like a diary entry. I sympathise, I just wish there were some conclusion the piece could come to, some action I could take or insight I could gleam. But nope, its just another soul trapped in the torment matrix.
(to be clear, there is good reason there is no call to action: a big part of the horror of all this is that there is no clear thing anyone can do)
I feel like people are trapping themselves with LLM use patterns. As OP has noted it's a pretty damn good self tutor tool, it's also great for offloading tasks you don't care about. But it will be devastating when people use it to do the important things foe them instead.
Potential bright side is that those people that will put in actual work and learn how to use AI as a complement and not replacement should have sizeable competitive advantage. LLMs are still dumb af in many ways.
The problem is that it's really hard for the average person (including myself) to understand which problems I should and shouldn't care about. Often the most trivial of tasks end up becoming inspirations to me, and I have some of the best ideas when I'm sweeping the floor.
It's almost as if the nature of rote or grunt work has a hidden beauty that provides negative space for creativity to flourish. And being forced to do those things without a choice is a great simplifier in life to maintain that contrast.
I was wondering about this the other day
Thank God OpenAI didn't come up with Docker. Otherwise we would be drinking water from bottles rhat are docker enabled
I’m being reminded of “journey before destination”, quite a bit since AI has come out.
Looking at people who are happiest leveraging it, it’s people who want to build something.
The people who are maximizing its utility are those with (deep) pre existing expertise.
For people who want to understand and think deeply about things, it’s … less useful. It can be used to create prototypes, or explain terms.
This is the path that is most precarious. Its utility is entirely dependent on being self motivated learners.
As the article showed, at some point people who want to just build, and don’t care about the learning, will just use ChatGPT to build their homework.
The situation seems like we are consuming the seed stock of expertise.
If evidence turns up of students and new entrants to the labour market finding ways to grow, learn skills or just find employment, the picture would clear up.
Not worth reading. A lot of whining about ChatGPT and lazy Oxford students, not enough substance, only "wisdoms" like "But mirrors can be dangerous if we stay for too long too charmed by our own reflection".
Classic slavic attitude (judging by your username). Worst software colleagues I ever had (you always have to be like binary with them: either you are right or wrong, either you provide something “valuable” or not… never something in between, never sympathy)
I'm glad I finished my Ph.D. before any of this stuff hit the market. It sounds like it would have been a substantially different experience, and I suspect I might not have finished it with as solid a grasp on my thesis subject as I have.
Her comment about the Mathematica notebook really hits home. As a scientist, not a software engineer, who uses code as a tool to speed up more critical parts of my work, I was thrilled to have the chance to learn programming skills on the job. I could justify it because the time saved by the fruits of it outran the time I spent learning it. I can't do that anymore. I'm not building software, just scripting, so even the cheapest models today can give me suitable code quicker than I could write it myself. So I can no longer justify learning new programming skills, at least on work time. I'm glad I had a few years to pick up the basics before these tools became available.
It’s the same at work. You don’t ask your colleagues for help anymore, you ask the agent. You can clone all the necessary repos, link to confluence/figma/drive and then ask.
It sucks. I don’t care about being efficient (that’s something the executives care about). I care about how I live my life, and at work that means learning from others (and offer the little knowledge I have to others)… if it takes 2 engineers 2h to solve a bug (vs 20 min for 1 engineer and an agent) then thats great for me. I get the chance to exercise my brain, solve a bug, learn from other engineer and have a bond with that individual. If I solve the bug in 20min with an agent, well, that’s something my boss would love, but I personally would have gained nothing… and would have to jump to solve more bugs alone in the remaining 100 minutes (which feels like running a marathon)
My anecdote is that most of my coworkers across multiple jobs didn't like to explain things. I think most engineers don't like being interrupted for questions and I would have appreciated a way to get answers without bothering them.
At the same time, there's a type of person who loves to explain things. They like to give presentations and are extra helpful when engaged. I'm pretty sure you can still ask them questions.
I don't like being interrupted, but I'm happy to answer a question when I'm at a good stopping/breaking point with whatever I'm working on.
Problem is that the type of person that leans towards collaboration and likes to explain things may be getting burned out right now or may be among the first to be replaced.
OTOH, it's great for the type of person like me, who likes answering questions but is rarely asking them. Like, I don't want to interrupt my colleagues for something I can figure out on my own in comparable amount of time/effort. AI is now scaling this to questions previously way too difficult to answer on my own.
Anecdata: I am that person, loves explaining things, love seeing my colleagues learn things and grow. Needless to say, I no longer hold a job in the industry.
I maybe liked to explain things, but never liked to ask other people, always liked to figure things by myself. Maybe it was the experience from school and university with a lot of teachers that didn't know what they were talking about, which made me not trust other people.
I'm the same as you. I hate asking other people, unless it's for a reference or material to learn. I like to grind to understand myself, and then answer questions and explain things via presentations or a drawing board. I never wonder why I'm like that, tho
That said, hardly anyone asks for explanations now. In my team people default to a chatbot for work questions almost exclusively, and in private life for hobbies I share with other people, they increasingly go for chatbots too.
Turns out, not everyone seems to want a human connection.
Oh, we do want a human connection. But we also want good information, and want to actually know this (i.e. hobby is something to do, not merely a pretext to socialize). The overlap between the two is low.
Efficiency in the short-term also leads to community degradation in the long term. And by efficiency I mean efficiency past the point of diminishing returns. Compiling code by hand would suck, but not writing code at all also sucks.
There is no formula to determine the right level of it, but it's one of those things that's obvious when you see it... like not eating.
I'm surprised by this sudden "I love interacting with humans" surge. It's not something I associate with hackers.
Neither do I. But what I suspect is that the unpleasantness of too much forced social interaction, which dominated discussions in times past, was also associated with the begrudgingly-acknowledged good it did, even for us introverts. Hey, I'm like that too. But AI has taken that away for many people, who are now realizing that occasionally eating those dark greens actually did some good and now they're gone, life is actually worse.
This stereotype of "hackers" feels very rooted in the late-90s/early-2000s. Ever since mid-2000s the prominent "hackers" have been collaborating in the open and chasing VC funding...
Not every hacker fits the stereotype of "guy typing fast in a basement filled with screens and terminals".
People are people, all different, some hackers enjoy the solitude of doing their project by themselves with no human interaction whatsoever, others enjoy and thrive with a collaborative setting, spitting ideas, showing code, and getting feedback to iterate on and improve them.
I've belonged to the latter camp most of my life even though I acted like the first when I was learning how to program (and on my first job); I'm not a stereotypical extrovert but love nerding about stuff together with someone: coding an idea, making music, working on wood, for me those are all more enjoyable with another human sharing the same passion.
Hackers are not anti-social, we're just a subculture. Hacking itself is ultimately a community activity. You want others to appreciate this cool thing you've made, or share tips about it, or just in general provide people connection. People buy plane tickets and conference tickets, pick their university/major, accept materially non-optimal job offers, just to seek out other members of this culture.
And for programmers that aren't hackers, they just appreciate the practical upsides of having mature humans that can take accountability for decisions that affect your work as coworkers.
> Hacking itself is ultimately a community activity. You want others to appreciate this cool thing you've made, or share tips about it, or just in general provide people connection.
It is, but it is largely also single-player.
There's a wide gap between "introvert" and "no interaction ever".
hacking something alone is fun, but maintaining something that’s being hacked by someone else without coordination has always been a shitty experience.
Even those that have less bandwidth for human interaction value being in the loop.
Presumably the people who loved interacting with humans were too busy interacting to show up on forums. And now that interactions are being curtailed they do end up here.
I've discovered that I either want to have no interaction with people, or I want to have real human interactions. I don't want the shitty half measure where my manager sends me whatever Claude sent them, or gives me instructions like I was a prompt ("do X now").
There are people working in IT or CS who aren't hackers.
I was mainly referring to comments on Hacker News.
perhaps consider what motivates a person to read comments on a forum like hacker news. Or indeed, post on such a forum.
There may be reasons other than self promotion
I feel there are workplaces and career choices where those people go who want social contact and fulfil personal utopian ambitions, not necessarily having profit in mind, and jobs where you go to make money efficiently. And these days, the divide between those has gotten very large to the point many jobs have no personal contact or room for personal growth, interests or artistic skill anymore. It feels the sane middle ground for sane people who are looking for a sane balance has been or is being hollowed out.
And when that workday is done, everybody goes home by themselves to scroll some more.
Do you apply this logic to all efficiency gains or just ones that weren’t commonplace when you entered the industry?
For instance, package managers didn’t always exist. Would you prefer it if we got rid of them? Instead of 30 seconds running `uv add whatever` collaborating with your package manager, you could spend 30 days collaborating with other human beings to build the functionality yourself.
I think the key question to answer for all of us - Where would you be in 5 years? :-)
In capitalism companies do things what is good for the company.
You can try convincing them that doing the same job in 2hrs is better for the company but they may not listen.
Like in outsourcing when they say the Indian worker does the same job for a much smaller salary and they don't care about the quality of the code they produce only about hitting the milestones. If the code is bad then they hire some other cheap outsourced workers to fix it, etc.
The market does not reward good code, the market rewards low price and speed of delivery.
Time and time again, I see the same complaints. A new technology comes and it removes bottlenecks. The ideal person should be ecstatic with new capabilities being unlocked. But they are complaining. Why?
They don't trust their peers to use it properly and externalise the concerns to the technology itself.
> The problem is that the little chatbot in our pockets is making us more isolated and in so, it’s making us stray further from this humble goal: to learn
I don't understand the author here, ChatGPT (as admitted by herself) does help people learn. Far more than what was possible before. Her only complaint is that society is not able to leverage ChatGPT to further the spirit of physics, but that's on her on peers and not on the technology.
As usual, society adapts given enough time.
> In particular, it would be pretty sad to destroy one of the few things that makes our life meaningful, not the fancy prize or the fancy position, but simply having a community to share and discuss and learn.
Exactly. The problem is a weird one though, because there are at least independent forces that are working together here:
(1) The people that are inventing AI or those that are pro-AI are typically those that have been conditioned to enjoy the "information rush" and place personal superificial enjoyment over human interaction. Lots of nerds/geeks simply love the idea of having a playground devoid of the necessity to actual interact outside their microcosm and so they think AI is the best thing since sliced bread. They're often people who were taunted in school or at least partially ostracized early in life because of their intelligence and take emotional comfort in tech.
(2) The perverted incentives of academia reward publishing and in general the incentives of capitalism create an arms race to be as efficient as possible at the cost of goods like "community" that only accrue vague benefits often outside the capitalistic system. So people tend to use AI.
(3) Very advanced technology itself weakens and even destroys smaller communities whose members then seek out technology as a superificial salve to ease what they consciously or subconsciously feel as something missing.
AI was never great. Knowledge as an end in itself is not a good thing and we should seek knowledge to better ourselves in a myriad of ways. But today it's just sought out for material and capitalistic advancement, which prevents our using it to actually become wiser and enjoy the basics of life.
The only people who defend it are those who are already addicted to it, or those who don't want to experience the cognitive dissonance of the truth of its harmful nature.
I somewhat agree with the article on the narrative level.
Lately, my thoughts on our LLM future darkened a tad, too.
I find the parallel between LLMs and Sophons from the 3-Body Problem novels increasingly accurate. That is, specifically, in terms of their effect of stalling human progress. Sure, anybody can opt out of using LLMs (unlike with Sophons), so the restriction on progress seems kind of "soft" in this sense, but the chances of someone choosing to do so aren't too high, IMO. It is so much easier to avoid forcing yourself to learn or explore hard stuff and just ask the model instead.
While I tend to support this in a field such as SW Engineering (if only the Standards and Quality Control aspects advanced as fast as raw code generation...), I share the same fears as Tao and others regarding Mathematics (and not just Mathematics; Computer Science, which ultimately underlies SWE, is on the spotlight as well).
>It is so much easier to avoid forcing yourself to learn or explore hard stuff and just ask the model instead.
What exactly is hard stuff though? The concept, the implementation, the grokking? For a person with the right mindset now is easier to get to the Eureka moment, because it is fairly cheap to try all the stuff that doesn't work.
Society benefits when in person collaboration builds understanding and creates new ideas and approaches /in both parties/.
Since now the former potential collaborators are feeding on more or less the same generated content, it's hard to see how AI will replace that.
And that's not the mention the further alienation it causes, as heavily alluded to in the article.
Reads like a diary entry. I sympathise, I just wish there were some conclusion the piece could come to, some action I could take or insight I could gleam. But nope, its just another soul trapped in the torment matrix.
(to be clear, there is good reason there is no call to action: a big part of the horror of all this is that there is no clear thing anyone can do)
I feel like people are trapping themselves with LLM use patterns. As OP has noted it's a pretty damn good self tutor tool, it's also great for offloading tasks you don't care about. But it will be devastating when people use it to do the important things foe them instead.
Potential bright side is that those people that will put in actual work and learn how to use AI as a complement and not replacement should have sizeable competitive advantage. LLMs are still dumb af in many ways.
The problem is that it's really hard for the average person (including myself) to understand which problems I should and shouldn't care about. Often the most trivial of tasks end up becoming inspirations to me, and I have some of the best ideas when I'm sweeping the floor.
It's almost as if the nature of rote or grunt work has a hidden beauty that provides negative space for creativity to flourish. And being forced to do those things without a choice is a great simplifier in life to maintain that contrast.
Beautiful writing style. So light, still eloquent.
[delayed]
I was wondering about this the other day Thank God OpenAI didn't come up with Docker. Otherwise we would be drinking water from bottles rhat are docker enabled
I’m being reminded of “journey before destination”, quite a bit since AI has come out.
Looking at people who are happiest leveraging it, it’s people who want to build something.
The people who are maximizing its utility are those with (deep) pre existing expertise.
For people who want to understand and think deeply about things, it’s … less useful. It can be used to create prototypes, or explain terms.
This is the path that is most precarious. Its utility is entirely dependent on being self motivated learners.
As the article showed, at some point people who want to just build, and don’t care about the learning, will just use ChatGPT to build their homework.
The situation seems like we are consuming the seed stock of expertise.
If evidence turns up of students and new entrants to the labour market finding ways to grow, learn skills or just find employment, the picture would clear up.
Not worth reading. A lot of whining about ChatGPT and lazy Oxford students, not enough substance, only "wisdoms" like "But mirrors can be dangerous if we stay for too long too charmed by our own reflection".
Classic slavic attitude (judging by your username). Worst software colleagues I ever had (you always have to be like binary with them: either you are right or wrong, either you provide something “valuable” or not… never something in between, never sympathy)