The breadth of BuzzFeed CEO Jonah Peretti's AI aspirations just got a lot clearer — and if he has it his way, AI use at the viral publisher won't be limited to time-killing quizzes and bottom-tier travel guides.
"BuzzFeed has always lived at the intersection of technology and creativity. And recent developments in artificial intelligence represent an opportunity to take this convergence to the next level," Peretti told eager investors at the company's Investor Day last week. "We view AI as an exciting new creativity tool, one that humans can harness to open up new avenues for imagination, storytelling and entertainment and explore new premium product offerings that allow us to innovate and collaborate with our clients and partners on a new frontier in media."
"Over the next few years, generative AI will replace the majority of static content, and audiences will begin to expect all content to be curated and dynamic with embedded intelligence," he continued. "AI will lead to new formats that are more gamified, more personalized, and more interactive."
First of all, we must say: "embedded intelligence" has got to be the most pretentious — not to mention vague — way to refer to integrated AI tools, and it's also not like we were expecting much from Peretti after he chose to kill the award-winning news division of the organization that was BuzzFeed News.
But even with the bar in the basement, this is disappointing. Partly because whatever Peretti's pitching sounds like an AI mad libs purgatory, and partly because it very much seems like the CEO is about to do exactly what, just a few short months ago, he explicitly said that he wouldn't: deploy the tech across BuzzFeed brands for what he derisively referred to as "cost savings and spamming out a bunch of SEO articles that are lower quality than what a journalist could do, but a tenth of the cost."
"Operationally, from an efficiency perspective, AI enables rapid prototyping of new content formats without the need to add fixed costs," said Martin, assuring that the company has "seen that audience engagement is multiples higher" with the "introduction of new AI power formats like chatbots and games" when compared to the site's static content.
"Instead of generating 10 ideas in a minute, AI can generate hundreds of ideas in a second. For example, our team members can talk directly to an AI chatbot to progress their creative ideas," Martin continued. "And earlier this year, our product team rolled out an exciting new feature that leverages AI to automatically suggest SEO headlines for our article based on the other headlines. And we have several more developments being built into our CMS to make content creation significantly easier."
"From a data and insights perspective," she added, "AI will continue to help us level up our existing first-party data offering for clients."
And speaking of data, user data certainly seems to be at the heart of BuzzFeed's new AI pitch. Throughout the presentation, BuzzFeed's featured execs fixated on AI as a means of delivering "hyper-personalized" content, which appears to mean putting readers, writers and creators, and advertisers in an endless data loop.
"Using the generative power of large language model AI, we can create hyper-personalized content at scale... that is far more engaging to our audience than static media, from chatbots to games to personalized articles and then we can feed the engagement data from that content back into the data pipeline," Jess Probus, senior vice president of BuzzFeed Editorial, told investors.
"I would characterize our application of AI across two buckets," she continued, "AI native content and AI as a copilot."
In other words: like Insider's employees have been encouraged to do, BuzzFeed's still-standing staffers will be expected to use AI to meet productivity goals. Even more depressingly, however, it sounds like they'll be using AI to generate an endless stream of seamless, hyper-targeted advertisements from scratch. Honestly, it's a little dystopian.
On that note, did we mention that BuzzFeed also says it'll use AI to tap into the "authentic voice" of various racial groups as a way for corporate clients to cash in on extra sales?
"In order to help [Sprite] reach a multicultural audience, the campaign ran across Complex and Hot Ones as well as BuzzFeed's Black Identity brand," said Andrew Guendjoian, Buzzfeed's VP of client partnership, adding that "we see creators, AI, and [cultural] moments as intersectional."
For those keeping score, this is hardly Peretti's first bizarre foray into racial politics. Way back in 2002, his first flirtation with viral fame was an incredibly uncomfortable website called Black People Love Us, where he and his sister Chelsea Peretti wrote insensitive and seemingly fictional testimonials about race ("Sally and Johnny give me ample opportunities to translate rap lyrics, reggae songs, and/or street slang!" read one.)
Of course, the regular AI caveats apply. Chatbots are notorious fabricators, so any text that they generate requires intense human scrutiny. In the presentation, the execs laid out their desire to distribute AI content "at scale." Given the size of BuzzFeed already, that's a lot of chatbot fact-checking to manage, especially if they plan to use the bots to rewrite and update past content as well, like CNET has reportedly done.
AI-generated material also presents plagiarism and copyright concerns, and as BuzzFeed itself has already demonstrated, AI-generated content is often relentlessly dull. And elsewhere, replacing and remixing static content also begs questions of ownership — if AI is constantly rewriting human-produced content, whether for personalization or SEO purposes, what stake does the original writer really have in their work? (If, moving forward, writers continue to be involved in the process in a real way at all, that is.)
It's all pretty bleak — and we're not the only ones with some thoughts on the matter.
"When news that BuzzFeed was deploying AI first broke, Peretti insisted that it would be used just to generate content like quizzes," tweeted tech columnist Brian Merchant, but now "it's becoming clearer now the plans went far beyond that, to an assumption that AI output could replace human work."
In any case, we wish Peretti the best of luck with his plagiarized, choose-your-own-adventure content farm, where Walmart and Burger King will use hyper-personalized, AI-generated-but-definitely-authentic advertisements to drive sales across racial demographics. Cheers.