It was dying well before AI and while a knee happened coincidentally to chatgpt release, I don't think you remember correctly what gpt answer looked like for programmers questions back then. It was much later where it started assisting development.

raincole 4 hours ago | parent | next [-]

It doesn't matter. Even if it were the most friendly community ever, as long as their core value were the same - answering programming questions that with clear answers - it would have been replaced by today's AI completely. It's irrelevant today how bad GPT3.5 once was.

▲

johnfn 4 hours ago | parent | prev [-]

I don't understand the motivation to reattribute the death of SO to anything other than AI, which it unquestionably was. GPT-3.5 was able to answer some small chunk of my programming questions (like 10-20% of them), which seems clearly reflected in the graph.

▲

dspillett 4 hours ago | parent | next [-]

> I don't understand the motivation to reattribute the death of SO to anything other than AI

A lot of people are rather bitter that they couldn't get the response they wanted from questions in the format they wanted to ask them, and would rather think that this slight against them is part of what killed the site. If it is all due to AI then they can't tell the tale of how SO rued the day its community deemed them off-topic :)

▲

kstrauser 4 hours ago | parent | prev [-]

Traffic started dropping precipitously in summer ‘20. There were no widely known LLMs then, let alone ones used outside labs. ChatGPT went public in November 2022, well into the decline.

While AI was probably the last straw, it only hastened the collapse that was already underway.

▲

Jtarii 4 hours ago | parent | next [-]

So you think there is a viable business around technical Q&A even with GPT 6 Astra or whatever existing?

LLMs so obviously make any type of the business model of SO unviable.

Attributing the death to mismanagement really does just seem like motivated reasoning due to hating SO.

	▲	kstrauser 4 hours ago \| parent [-]
		If there is, I’m not creative enough to imagine it. I don’t contend that SO could survive AI. You’ll never hear me arguing that. However, that graph shows that it was already dying before AI became widely available. AI 100% certainly kicked it while it was down. It just cannot have been the cause of the downturn because the downturn was in full swing before AI could’ve had any effect at all.

▲

cwnyth 4 hours ago | parent | prev | next [-]

I think you're forgetting of a certain major event that happened in 2020. An event that initially caused a lot of free time around the world and kept people indoors...

▲

kstrauser 4 hours ago | parent [-]

You mean, the kind that made them more likely to sit around asking and answering questions because there wasn't anything else do to, and which caused an initial spike of traffic?

▲

cwnyth 3 hours ago | parent [-]

I think it's pretty clear that was the cause of the 2020 spike and fall. Once people started going back to work (summer 2020 for us), questions dropped back to pre-2020 levels.

	▲	kstrauser 3 hours ago \| parent [-]
		Sure, but it rapidly fell before its pre-COVID levels and then kept dropping rapidly for the next year or so, and then AI really kicked in.

▲

bsder 4 hours ago | parent | prev [-]

Agreed.

StackOverflow was dead man walking as soon as some HR idiot at Google weaponized Fake Internet Points via quarterly reviews needing "social contributions".

Before that, the ignorant punters would leave these kinds of sites alone. Since there was nothing to be gained, you either had to be an altruist or deranged to waste your time on answering questions on the site.

After Google weaponized Fake Internet Points and everybody copied Google, a zillion software programmers everywhere in the world suddenly wanted the shortest path to earn lots of Fake Internet Points rather than actually answering relevant questions.

THAT was the death of StackOverflow--everything else has been the denouement.