I reached out for help. The standard response from everyone fell between ignoring me or telling me I need to "do it myself".
You know who did help me? Random strangers on Stack Overflow. People who had nothing to gain from helping me other than just trying to help someone in need.
This part makes the entire post sound like a parody. Maybe Jeff Atwood has a wicked, self-deprecating sense of humor?
Yeah no, ask for help on Stack Overflow and you were were more likely to be met with castigation and snide remarks. It was such a breath of fresh air when ChatGPT let us not have to deal with that community anymore.
I dunno man. In it's heyday, SO was pretty fucking awesome. IMHO. Just a massive amount of good information concentrated. And some bad, but a damn sight better than random Google searches even back then.
ChatGPT also isn’t going to tell you when your idea is a bad one and will happily assist you in your (unknowing) efforts to brick your computer. So there’s that.
You are very much talking about late era StackOverflow there.
What marked it out originally (and was visible in some other less-travelled StackExchange sites rather later) was a culture of holding replies to a sort of account in the reply comments; people urging each other to make their replies better-sourced, less hollow, more responsive.
The goal of being non-duplicative alas ultimately is always going to lead to a culture that polices itself too hard. I am an LLM cynic but I think it’s possible that LLMs could actually help steer that process more gently, guiding submitters to likely answers but also helping them fine-tune a more specific, narrow, never-asked question.
On the GrapheneOS forum, I enjoy answering the questions in a post of user who just set up their phone.
AI does not have as good knowledge of the OS as frequent forum contributors. I'm sure plenty of people get by with AI help for their questions and there's likely little questions that hasn't been already discussed, but I love to see new people using the OS and it doesn't take long to give them an exact answer. The Flarum software that the forum runs is very fast, as well.
Humans are social animals. It's true that not all of these interactions re positive, but so much of what drives us is not even related to acclaim or flattery, but merely interaction in a general sense. In a sense this is one of the true perils of LLM-style interaction (going back to Eliza, I guess) -- that we can not only potentially satisfy that drive mechanically, but do so in a way that's structurally produced by a commercial entity. Do you think that the models talking to you have your best interests at heart? They don't have any interest in you at all.
My personal rule as a recovering grammar school victim: don’t correct nits if you’re not going to add some cool trivia or at least provide some gentle, kind way to navigate the same confusion next time.
Less/fewer is probably a lost cause, and a little bit of why is this kind of useless correction. If those who correct can’t or won’t explain why one is correct over the other, then it simply isn’t.
Capitalism is optimized to produce products and services people want to buy which drives people wanting to work more hours. Which then drives up the costs of things that aren’t replicated endlessly like land and goods that are intentionally left scarce such as admission to elite institutions.
You can work dramatically less and still live well, but the sirens call is very seductive. Even our medical system is optimized to extract value instead of optimizing care. Food, clothes, transportation, housing, etc all offer real value at minimal cost and little pushing you in that direction.
The flaw with this argument is the assumption people who control the LLMs is the sole problem. Plenty of people use LLMs for personal gain at the cost of others.
At some point LLM users and proponents have to come to terms with the fact that this has all been a massive theft of labor, including donated labor and mutual aid in forums such as Stack Overflow.
Such a massive theft of labor will have negative consequences that we all have to live with.
There are many ways to steal labor, violating copyright is one such way, but not all copyright violations is labor theft, nor does labor theft require copyright to be violated.
The problem is that this source of free labor has dried up, precisely because of AI.
I've noticed that in "programming" forums, questions have devolved.
In the "before times", you'd get a mix of simple-medium-hard-extreme questions.
Google and StackOverflow made most of the medium and hard questions disappear.
Now with LLMs, all the higher difficulty levels except "beyond extreme" have vanished, leaving only the stupidly difficult questions that nobody can answer -- and hence doesn't contribute to collective knowledge, including the AI's knowledge.
Conversely, there is a new level of "below simple" questions by people that I dare say are too stupid to even attempt to become software developers. The type of questions that can could always have been answered with just Google, and are "beneath" both AIs and humans. These are worthless questions along the lines of "my computer not work, plz halp!?" followed by "I can't reboot it!" as if that makes any sense.
Again, that contributes nothing of value to the zeitgeist.
What's going to happen in 5 or 10 years from now, when all of the pre-AI knowledge[1] is stale and out of date, making it irrelevant, but there's no new knowledge being publicly produced to train on?
Will programming languages, tools, and frameworks ossify because new languages are "post-AI" and hence untrainable due to a lack of StackOverflow questions and answer?
Will AI labs build their own private "AI members only" version of StackOverflow, building up their internal knowledgebases for training, sharing this with nobody else?
Will humans be paid to solve problems for the sole purpose of generating AI training materials? Is this the future of knowledge work!?
Stack overflow was one of the best ways to learn and mature as an engineer “back in the day”. The
meme is that devs would copy and paste and hope it worked. The reality was more like you had an obscure bug, you wrote a detailed and cohesive question, and you connected with people who were more knowledgable and would help you.
It also sharpened your writing and thinking. By responding to questions you solidified your understanding of topics.
It got a lot of hate for the moderation but I think that was mainly in huge communities where literally “dumb” questions got bombarded every minute. Smaller tech like Docker when it first arrived on the scene had communities that learned together.
Spent a lot of time there when I was a few years out of school and still trying to figure out what it meant to be a software eng. It’s sad to see it become a ghost town and I think it’s a net negative for young engineers to have an all knowing entity spit out every answer.
This part makes the entire post sound like a parody. Maybe Jeff Atwood has a wicked, self-deprecating sense of humor?
Yeah no, ask for help on Stack Overflow and you were were more likely to be met with castigation and snide remarks. It was such a breath of fresh air when ChatGPT let us not have to deal with that community anymore.
Plus Stack Overflow was gamified. "Nothing to gain" except 12 different artifacts showing how awesome you are.
I dunno man. In it's heyday, SO was pretty fucking awesome. IMHO. Just a massive amount of good information concentrated. And some bad, but a damn sight better than random Google searches even back then.
ChatGPT also isn’t going to tell you when your idea is a bad one and will happily assist you in your (unknowing) efforts to brick your computer. So there’s that.
You are very much talking about late era StackOverflow there.
What marked it out originally (and was visible in some other less-travelled StackExchange sites rather later) was a culture of holding replies to a sort of account in the reply comments; people urging each other to make their replies better-sourced, less hollow, more responsive.
The goal of being non-duplicative alas ultimately is always going to lead to a culture that polices itself too hard. I am an LLM cynic but I think it’s possible that LLMs could actually help steer that process more gently, guiding submitters to likely answers but also helping them fine-tune a more specific, narrow, never-asked question.
I found SO pretty helpful and approachable in the day. Not perfect, but that’s true of anything involving people.
On the GrapheneOS forum, I enjoy answering the questions in a post of user who just set up their phone.
AI does not have as good knowledge of the OS as frequent forum contributors. I'm sure plenty of people get by with AI help for their questions and there's likely little questions that hasn't been already discussed, but I love to see new people using the OS and it doesn't take long to give them an exact answer. The Flarum software that the forum runs is very fast, as well.
Humans are social animals. It's true that not all of these interactions re positive, but so much of what drives us is not even related to acclaim or flattery, but merely interaction in a general sense. In a sense this is one of the true perils of LLM-style interaction (going back to Eliza, I guess) -- that we can not only potentially satisfy that drive mechanically, but do so in a way that's structurally produced by a commercial entity. Do you think that the models talking to you have your best interests at heart? They don't have any interest in you at all.
In this sense, running (and heavily tuning) a local LLM could be seen as a form of self-authorship.
Not that I find them all that useful, still.
We had better have even more time to help each other as a result of LLMs, or what are we even doing here?
I realize that everything seems to be about increasing the pace of life, but this is another "chance" for humanity to figure things out.
No increases in productivity ever led to less working hours, only organized action. We work more hours than miedeval serfs.
Fewer working hours.
My personal rule as a recovering grammar school victim: don’t correct nits if you’re not going to add some cool trivia or at least provide some gentle, kind way to navigate the same confusion next time.
Less/fewer is probably a lost cause, and a little bit of why is this kind of useless correction. If those who correct can’t or won’t explain why one is correct over the other, then it simply isn’t.
Serfs had few incentives to work more hours.
Capitalism is optimized to produce products and services people want to buy which drives people wanting to work more hours. Which then drives up the costs of things that aren’t replicated endlessly like land and goods that are intentionally left scarce such as admission to elite institutions.
You can work dramatically less and still live well, but the sirens call is very seductive. Even our medical system is optimized to extract value instead of optimizing care. Food, clothes, transportation, housing, etc all offer real value at minimal cost and little pushing you in that direction.
The people who control the LLMs would rather be 0.1% richer than have any/every other person on the planet feel better in any way.
The flaw with this argument is the assumption people who control the LLMs is the sole problem. Plenty of people use LLMs for personal gain at the cost of others.
Every scope of help commoditised. Help for late night drop. Uber ! Need for friendly food door dash !
At some point LLM users and proponents have to come to terms with the fact that this has all been a massive theft of labor, including donated labor and mutual aid in forums such as Stack Overflow.
Such a massive theft of labor will have negative consequences that we all have to live with.
https://news.ycombinator.com/item?id=49622042
You can't steal something that's infinitely duplicatable. Also, AI training is fair use, not that I believe it should have to be.
Copyright violation ≠ theft of labor.
There are many ways to steal labor, violating copyright is one such way, but not all copyright violations is labor theft, nor does labor theft require copyright to be violated.
The problem is that this source of free labor has dried up, precisely because of AI.
I've noticed that in "programming" forums, questions have devolved.
In the "before times", you'd get a mix of simple-medium-hard-extreme questions.
Google and StackOverflow made most of the medium and hard questions disappear.
Now with LLMs, all the higher difficulty levels except "beyond extreme" have vanished, leaving only the stupidly difficult questions that nobody can answer -- and hence doesn't contribute to collective knowledge, including the AI's knowledge.
Conversely, there is a new level of "below simple" questions by people that I dare say are too stupid to even attempt to become software developers. The type of questions that can could always have been answered with just Google, and are "beneath" both AIs and humans. These are worthless questions along the lines of "my computer not work, plz halp!?" followed by "I can't reboot it!" as if that makes any sense.
Again, that contributes nothing of value to the zeitgeist.
What's going to happen in 5 or 10 years from now, when all of the pre-AI knowledge[1] is stale and out of date, making it irrelevant, but there's no new knowledge being publicly produced to train on?
Will programming languages, tools, and frameworks ossify because new languages are "post-AI" and hence untrainable due to a lack of StackOverflow questions and answer?
Will AI labs build their own private "AI members only" version of StackOverflow, building up their internal knowledgebases for training, sharing this with nobody else?
Will humans be paid to solve problems for the sole purpose of generating AI training materials? Is this the future of knowledge work!?
[1] https://en.wikipedia.org/wiki/Low-background_steel
Stack overflow was one of the best ways to learn and mature as an engineer “back in the day”. The meme is that devs would copy and paste and hope it worked. The reality was more like you had an obscure bug, you wrote a detailed and cohesive question, and you connected with people who were more knowledgable and would help you.
It also sharpened your writing and thinking. By responding to questions you solidified your understanding of topics.
It got a lot of hate for the moderation but I think that was mainly in huge communities where literally “dumb” questions got bombarded every minute. Smaller tech like Docker when it first arrived on the scene had communities that learned together.
Spent a lot of time there when I was a few years out of school and still trying to figure out what it meant to be a software eng. It’s sad to see it become a ghost town and I think it’s a net negative for young engineers to have an all knowing entity spit out every answer.