Stack Overflow 論壇因 AI 式微，但公司卻靠 AI 轉虧為盈

Stack Overflow’s forum is dead thanks to AI, but the company’s still kicking... thanks to AI

The platform is raking in millions of dollars in revenue, with AI an ironic new source of revenue.

When Elon Musk described Stack Overflow’s plight as “death by LLM” in July 2023, he wasn’t exaggerating.

Having been the go-to resource for developers looking for technical help for a long time, Stack Overflow neared the peak of its powers during the pandemic, with coders seeking the evergreen information on the company’s popular Q&A forum. But amid a wave of powerful code-writing AI assistants like ChatGPT, Cursor, Claude, Google’s Gemini, and Microsoft’s Copilot, traffic to the site has plummeted.

Last month, Stack Overflow recorded just 6,866 questions — roughly equal to the typical volume when the site first launched back in 2008.

But while Stack Overflow the Q&A forum looks dead, Stack Overflow the company looks to be limping along.

Unlike Chegg and other knowledge hubs that have fallen victim to generative AI, Stack Overflow has found a way to monetize its enormous back catalog of content. Indeed, even with engagement falling off a cliff since ChatGPT’s 2022 debut, the company’s annual revenue has roughly doubled to $115 million. Losses have slimmed, too, from $84 million in FY2023 to $22 million as of the last fiscal year, as desperate cost-cutting efforts, including mass layoffs, helped boost the bottom line.

Once dependent on ads across its buzzy forum, Stack Overflow now primarily makes money from enterprise solutions like “Stack Internal,” which provides a generative-AI add-on powered by the millions of questions and answers on the site through the years. Stack Internal is now used by 25,000 companies around the world. It also licenses its data to AI companies, in a Reddit-like model — a platform that made more than $200 million from licensing user-generated content in 2024.

Put simply, Stack Overflow’s new niche is the trust built by its old community and their expertise. In the words of CEO Prashanth Chandrasekar last December:

“...when we saw the questions decline in early 2023, what we realized is that pretty much all those declines were with very simple questions. The complex questions still get asked on Stack because there’s no other place. If the LLMs are only as good as the data, which is typically human curated, we’re one of the best places for that, if not the best for technology.”

Large language models want data about coding problems and how to solve them. Stack Overflow has a big digital warehouse full of that, but it’s increasingly aging, as queries move into private chat windows with LLM models... which need huge chunks of data to work. Stack Overflow has become a fascinating canary in tech’s new, circular coal mine.

More Tech

AI leaderboard maker LMArena hits $1.7 billion valuation

If you want to know who’s up and who’s down in the AI model world, look no further than LMArena’s leaderboard. The startup has just raised a $150 million series A fundraising round, with a valuation of $1.7 billion.

In seven months, LMArena has raised $250 million, according to TechCrunch.

The leaderboard started as a research project by cofounders Anastasios Angelopoulos and Wei-Lin Chiang when they were graduate students at UC Berkeley.

The public leaderboard — formerly known as “Chatbot Arena” — shows the results of human evaluations of AI models for various tasks. Users can rate which model did a better job on one task in a sort of blind taste test.

The leaderboard is a hotly contested proving grounds for new models, and the company occupies a powerful position in an industry that lacks independent, industry-standard evaluations.

LMArena lands $1.7B valuation four months after launching its product | TechCrunch

The leaderboard started as a research project by cofounders Anastasios Angelopoulos and Wei-Lin Chiang when they were graduate students at UC Berkeley.

The public leaderboard — formerly known as “Chatbot Arena” — shows the results of human evaluations of AI models for various tasks. Users can rate which model did a better job on one task in a sort of blind taste test.

The leaderboard is a hotly contested proving grounds for new models, and the company occupies a powerful position in an industry that lacks independent, industry-standard evaluations.

Uber jumps after unveiling Lucid robotaxi at CES

Uber shares jumped more than 5% after the company unveiled a production-intent robotaxi developed in partnership with Lucid and Nuro at the Consumer Electronics Show on Monday. The autonomous vehicle runs on Nvidia’s Drive AGX Thor computer. Nvidia itself announced a slate of autonomous hardware and software announcements at CES.

The companies said this fall that the San Francisco Bay Area will be the first market for the joint effort. The robotaxi is already being tested on public roads, with a commercial launch planned for later this year.

Uber + Lucid + Nvidia is just another example of the tangled web of partnerships in the autonomous driving space, where Nvidia is now becoming more and more prominent.

This is Uber's new robotaxi from Lucid and Nuro | TechCrunch

The companies said this fall that the San Francisco Bay Area will be the first market for the joint effort. The robotaxi is already being tested on public roads, with a commercial launch planned for later this year.

Uber + Lucid + Nvidia is just another example of the tangled web of partnerships in the autonomous driving space, where Nvidia is now becoming more and more prominent.

Meta delays international Ray-Ban Display expansion thanks to “unprecedented demand” and “extremely limited inventory”

Meta said today that it’s delaying the early 2026 international expansion of its Ray-Ban Display glasses because of “extremely limited inventory” and “unprecedented demand.” The company didn’t specify whether the issue was more supply or demand, but has previously insisted its smart glasses are a hit.

Waitlists for the smart glasses, which are controlled with a band you wear on your wrist, extend “well into 2026.”

“We’ll continue to focus on fulfilling orders in the US while we re-evaluate our approach to international availability,” the company wrote. Expansion had been planned for the UK, France, Italy, and Canada.

In order to buy the smart glasses, consumers must do an in-person product demo to ensure the tech is “properly fitted to you,” according to Meta. Demos in New York City are unavailable for the next few weeks, the company’s scheduling website shows. It also notes that “that due to high demand, the product may be sold out and unavailable for purchase after your demo.”

Nvidia’s autonomous tech gives other automakers a chance to take on Tesla

Nvidia made a number of autonomous vehicle announcements at CES yesterday that should have Tesla worried.

Report: Returns on Nvidia’s Omniverse business are underwhelming

Today in Las Vegas, Nvidia’s Jensen Huang will be taking the stage at the CES conference to show off his company’s latest innovations. If you’ve watched any of Huang’s impressive jargon-filled keynote speeches, you’ll notice that he gets extra animated when talking about “digital twins.”

Nvidia’s “Omniverse” platform allows companies to use Nvidia software and hardware to run large-scale simulations of factories and assembly lines.

The idea is that you can train your robots in Nvidia’s simulated environment, saving huge amounts of time and money. Huang is making a huge bet on this nascent product, and according to a new report, it isn’t exactly taking off.

The Information reports that much to Huang’s anger and frustration, the Omniverse offerings have yet to generate significant revenue.

According to the report, despite a long list of customers who Nvidia says are using Omniverse, some say the software is hard to use and fails to allow for simulating robot interactions with complex objects. The report also says Nvidia shuttered its Omniverse Cloud service in August of last year due to a lack of demand.

Nvidia has certainly successfully blazed trails into new categories before, but considering Huang’s enthusiasm for the Omniverse offerings, it is a rare stumble for the chip juggernaut.

Nvidia’s Big Ambitions to Solve Manufacturing Shows Slow Returns So Far

The idea is that you can train your robots in Nvidia’s simulated environment, saving huge amounts of time and money. Huang is making a huge bet on this nascent product, and according to a new report, it isn’t exactly taking off.

The Information reports that much to Huang’s anger and frustration, the Omniverse offerings have yet to generate significant revenue.

According to the report, despite a long list of customers who Nvidia says are using Omniverse, some say the software is hard to use and fails to allow for simulating robot interactions with complex objects. The report also says Nvidia shuttered its Omniverse Cloud service in August of last year due to a lack of demand.

Nvidia has certainly successfully blazed trails into new categories before, but considering Huang’s enthusiasm for the Omniverse offerings, it is a rare stumble for the chip juggernaut.

Latest Stories