After ChatGPT disruption, Stack Overflow lays off 28 percent of staff
arstechnica.com
After ChatGPT disruption, Stack Overflow lays off 28 percent of staff::The popular developer forum is still hunting for a "path to profitability."
After ChatGPT disruption, Stack Overflow lays off 28 percent of staff::The popular developer forum is still hunting for a "path to profitability."
Great. So once Stack Overflow is dead, where will ChatGPT get actual, correct answers from?
Comment Closed: Duplicate Post
See other comment about different company going out of business for totally different reason.
Perfectly toxic, as all stack overflow comments should be.
They also went out of business 10 years ago and the market has changed since then.
An actual problem to worry about too. I think there will always be people looking to contribute but as less people do AI may actually get dumber until they figure out how to train AI with AI
That won't work because machine learning doesn't actually understand what it says. It needs real human knowledge underlying it. It can't just learn things on its own out of nowhere.
But maybe if we sacrifice enough ecosystems we could get it to work and then ask it to solve all the climate problems we created to power it...
That sounds exactly like what an AI, that was trained by another AI, would say to assuage our fears of General Artificial Intelligence. Nice try.
US Robotics would like to give you the first robot for free. It's Three Laws safe! We swear!
Wait… am I an AI?
That's true for general purpose LLMs, but there are other contexts in which machine learning models acquire knowledge without continuous human input, e.g. AlphaZero.
Ok but how does it incorporate wisdom in it's output? Or discern truth from lie?
I mean, the AI can memorize the programming documentation, sweep different github repositories, and the programming itself is already learned behavior.
That's for programming. As for fault finding, that might get more challenging for the AI without stack overflow.
I’m skeptical that an LLM could answer questions as effectively just with documentation. A big part of the value in stack overflow and similar sites is that the answers provided come from people who have experience with a given technology and have some understanding of the pain points. Often times you can ask the wrong question and still get a useful answer because the context is enough for others to figure out what you might be confused by.
I’m not sure an LLM could do the same just given the docs, but it would be interesting to see how close it could get.
To add to this comment. Most of the questions and answers in stackoverflow stem from situations not covered by the documentation or when the documentation fails. LLMs don't have a way to learn about these issues and how to address them because they require actual implementations to assess/validate.
Its the same reason why git repositories would also fail to meet this need. Repositories only contain (typically working code) without much context on why changes were made or were needed. Technically githib issues or jira tickets could help cover the gaps of something like stackoverflow dissappearing, but would ultimately mean that the information could be locked behind paywalls or corporate systems.
I can also ask basic, repetitive questions framed exactly to my use case without getting yelled at that the question has been asked and answered before.
and then be provided with a solution of a 10 years old tech that is barely applicable today. But if you read all the answers you may find an up to date suggestion in the comments of a non-accepted answer.
Honestly this is not bad, if it solves your problem and it took less than 10 minutes of reading overall.
Plus you gain some understanding along the way, about why the other answers aren't going to solve your problem, which is also valuable.
I'm not against learning or understanding the nature of the issue so that I'm in position to form a correct solution. But I find it kinda funny that topics are locked with "duplicate", and the question linked has for example an accepted answer written in PHP5. I believe that SO should invent a way that such cases can be updated without the current "hacks". Now, even if somebody puts the effort to write an up-to-date answer, it will be at the end of answers and it will take a lot of time, if ever, to reach a higher position. Also the "accepted" answer will always be the old one.
Another issue is that in some cases, the accepted answer becomes a wiki. Edit upon edit by users with required reputation adding new up-to-date info. In the end the answer is nothing close to what the original answer was, but it keeps the originally acquired score. So it can reach to a case that you see a +100 upvoted answer, which in fact has only been proof-read by 2-3 people, the ones that are needed to accept the edited answer.
Then if you try to provide a modern solution, get yelled at because 'not everyone is using the latest version' even though the modern solution works on everything newer than about 8 years.
The answers on SO are from unpaid contributors.
As if that excuses their behavior on a forum dedicated to learning and solving problems?
Everyone gets annoyed at oft-repeated questions. But the answer is moderation, which is also unpaid.
Not vitriol and skullduggery.
I have never even considered asking a question for that reason.
This is the best summary I could come up with:
Stack Overflow used to be every developer's favorite site for coding help, but with the rise of generative AI like ChatGPT, chatbots can offer more specific help than a 5-year-old forum post ever could.
While no chatbot is 100 percent reliable, code has the unique ability to be instantly verified by just testing it in your IDE (integrated development environment), which makes it an ideal use case for chatbots.
Today, CEO Prashanth Chandrasekar announced Stack Overflow is laying off 28 percent of its staff.
Of course, the great irony of ChatGPT hurting Stack Overflow is that a great deal of the chatbot's development prowess comes from scraping sites like Stack Overflow.
OpenAI is working on web crawler controls for ChatGPT, which would let sites like Stack Overflow opt out of crawling.
As we've seen with chatbots convincing each other that you can "melt eggs," Chandrasekar has argued that sites like Stack Overflow are essential for chatbots, saying they need "to be trained on something that's progressing knowledge forward.
The original article contains 276 words, the summary contains 168 words. Saved 39%. I'm a bot and I'm open source!
Or, you know, fact check before you feed it to the AI. You don't tell a child 'go google it' to everything and hope it somehow works out, right? "But then AI could never be profitable!" Oh the irony.
Btw, would a common language model be possible?
I don't get all the hate and vitriol for StackOverflow. Sure, some people are assholes. Welcome to humanity. At least the system provides for voting to suppress the shit takes and general assholery.
SO combined with Google is usually enough to help me find an answer that either gives the context I need to make a solution or a straight up solution. If people are posting and expecting a super detailed, correct answer in a matter of hours, I think their expectations need adjustment.
I've posted very few questions and had decent responses for the majority of them. Is my experience uncommon?
But yeah, layoffs suck, and I hope they find a way to be profitable. Hell, if they do a Patreon-esque model where people can just throw money at them because they appreciate the service, I'd subscribe. (If a similar thing exists that I don't know about, please link)
Hate is often meme-based today. Now people who never even thought of that site are suddenly against it because it's being talked about.
At-least Chat GPT doesn't yell at me for asking a question
“Apologies for the oversight, here is the corrected version that includes what you were asking for…”
Angry nerds are a valuable untapped resource if you're willing to tolerate our impatience
28%? Why, that's 7 people!