Netflix reported its quarterly earnings on Tuesday, and revenue was $11.51 billion—slightly below Bloomberg’s forecast of $11.52 billion. Earnings per share were $5.87, below an estimate of $6.94. (All of this is per Yahoo! Finance). Not catastrophic, but not great. The company’s stock price went down by 5.6%.

But hey, how about all that AI?

In its shareholder letter, and then again in its earnings call, Netflix co-CEOs Ted Sarandos and Greg Peters touted coming generative AI implementation as an exciting new development across content generation, user experience, and in advertising. All the AI talk made it feel like it was 2023 again.

To illustrate how much AI seems to be on the CEOs’ minds, at 26 minutes and 34 seconds into the earnings call, Peters gives an answer in which he rattles off what seems to be an ad-hoc list of six challenges Netflix faces as a company, starting with creating TV and movies around the world. Implementing AI is the second item he mentions.

The shareholder letter kicked off the topic of AI with what is actually a pretty big truth bomb: “For many years now, ML and AI have been powering our title recommendations as well as production and promotion technology.” That’s truer than you might think. All the way back in 2008, Netflix found a genuinely fun way to publicize its use of machine learning: by asking the public to figure out why some people like Napoleon Dynamite and others don’t, with math.

Much has changed since those days, and now Netflix says it’s shoehorning AI features into Netflix wherever it can. For instance, the earnings letter says a beta test is ongoing that adds a “conversational search” function, helping people “discover the perfect title for that moment.” The goal here sounds easy enough to picture without speculating too wildly. I can picture saying “What’s a movie to watch with my mom on her 50th birthday?” into a Roku remote and getting a list of recs.

On the techical side, Netflix says there’s an effort afoot to use AI to “localize” (which basically means “translate”) promotional assets into other languages and regions, which sounds like it could theoretically allow relevant—but obscure and foreign—movies and TV that I might love to come to me, even without readily available artwork and summaries in my language. Theoretically anyway.

But most of all, Netflix says it will be “empowering creators with a broad set of GenAI tools to help them achieve their visions.” For instance, the letter boasts that AI de-aging was used in Happy Gilmore 2, and that the producers of Billionaires’ Bunker used Generative AI tools for concept art.

But how does Netflix regard Sora 2, the video generator that plunged the internet into copyright hell upon its release last month? Sarandos, for his part, is not telegraphing some kind of rapid slopification at Netflix—and he hastens to add that he’s not intimidated by Sora 2 currently. According to Sarandos, “it’s likely to have a lot more impact on UGC creators the most in the near term. In other words, AI content replacing viewing of existing user-generated content. That starts to make sense.”

It’s worth noting that Sarandos is already trying to convince influencers to leave YouTube. If you’re a YouTuber who hasn’t already agreed to a Netflix deal, and you’re worried that Sora 2 is coming to eat your lunch, Sarandos’ suggestion that your business model is more vulnerable than his could be part of an effort to win you over.

But Sarandos says AI “doesn’t automatically make you a great storyteller if you’re not.” For once a tech CEO talking about AI doesn’t sound like he’s sharpening his layoffs ax. “We’re confident that AI is going to help us and help our creative partners tell stories better, faster, and in new ways,” Sarandos said, adding, “We’re all in on that.”