With an influx of Meta alums, some OpenAI staffers worry it is adopting Meta's tactics, like using social media dynamics with Sora and a softening stance on ads
There will be ads in OpenAI products faster than even they expect. [image] Casey Newton / @caseynewton : Really the most Meta-like tactic on display from OAI lately is to release products that have obvious harms and say “oops!” two days later and add some minor safeguard [embedded post]
The Information
Context & Ripple Effects
OpenAI had already signaled it was considering advertising, while recruiting ad talent from Meta and Google, in its earlier exploration of an ads model. The new concern is that product and monetization choices are now being read internally through the lens of those hires.
The staffing tension follows OpenAI's effort to counter Meta recruiting, including a response to Meta's talent offers that framed departures as a serious competitive threat. That makes cultural influence—not just headcount—a consequential part of the competition.
First-order effects
OpenAI staff concerned about the Meta influx face a sharper internal conflict over whether Sora's engagement mechanics and a faster ad rollout fit the lab's product standards.
Advertisers and prospective commercial partners gain a clearer signal that OpenAI products may become monetizable sooner, even as the company must manage user and employee trust.
Second-order effects
Meta's consumer-AI playbook becomes a more immediate benchmark for rivals: engagement-oriented features and advertising may be treated as a route to fund broad AI distribution rather than a last resort.
If leading AI products increasingly rely on feeds, engagement loops, and advertising, the competitive center of gravity could shift from model access toward consumer distribution and retention.
The trade-off is durable: commercialization can finance reach, but repeated safety fixes after launch may erode the trust that differentiates AI products from conventional social platforms.
The trend: This is one data point in the commercialization of consumer AI, where model companies are adopting distribution and monetization practices pioneered by social platforms.
The claim at the end about optimizing for engagement seems much more important than the ads question IMO, legally / ethically / reputationally etc. https://x.com/...
This really helps explain why some of OpenAI's recent releases like Sora, ads, and commerce make a lot more sense now. About 1 in 5 OpenAI staffers come from Meta. Even with massive use of their models and API, about 70% of OpenAI's $13B revenue still comes straight from [image]
The recent report from The Information is truly fascinating. As OpenAI aggressively hires former Meta employees in pursuit of monetization, there are growing internal complaints that the company might be turning into “another Meta.” What's especially interesting is that OpenAI [i…
From ChatGPT's sycophancy crisis in Q1 to the hiring of Meta/Facebook vets and the launch of Sora app, some longtime OpenAI employees have gotten concerned about the company's Meta-fication. [image]
An epiphany for me about consumer perception of ads: when I worked at Skype, an exec gave a presentation about a series of focus groups he ran about inserting ads into the product. A non-trivial number of participants asked, “Skype is free, so doesn't it already include ads?” [im…
OpenAI has imported approximately 630 former Meta employees representing about 20% of its roughly 3,000 staff according to LinkedIn data analyzed by The Information, with OpenAI's Slack having a channel just for former Meta employees according to a current employee and many [imag…
As OpenAI grows, employees are getting concerned that it's starting to look too much like the social media giant Meta. Details incl: - a metafication survey - internal reactions to sora - the rise of ads w/ the dream team @KalleyHuang @erinkwoo https://www.theinformation.com/ ...
>be Sam “ads are icky” Altman >May 2023: “last resort, uniquely unsettling, literally gives me hives” >promise Congress my AI will never sell your soul >quietly hire 630 Meta alums, rename HR to “Human Re-targeting” >new Slack channel #ad-dungeon already has custom Zucc
The idea that OpenAI will build products for erotic text porn, deepfakes, military applications and partner with ICE but then draw the line at showing you ads for toothbrushes is hilarious. A straight up joke. — There will be ads in OpenAI products faster than even they expect…
Really the most Meta-like tactic on display from OAI lately is to release products that have obvious harms and say “oops!” two days later and add some minor safeguard [embedded post]