Tech
Sora Space Dog 10FPS
PIVOT
TO VIDEO
(Courtesy OpenAI)

OpenAI’s Sora 2 started off scorching hot. Things have slowed down since.

The app racked up 1 million downloads in its first five days, despite being iOS-only and requiring an invite. But for all of December, it was downloaded just over 3 million times between iOS and Android.

Ryan Broderick, Adam Bumas

OpenAI’s stand-alone app for the AI video generator Sora 2 launched on iOS in late September. Despite initially requiring an invite to sign up, it was an instant success, reaching No. 1 overall on the iOS App Store’s charts after just four days. A day later, it cracked 1 million total downloads, which was faster than ChatGPT.

There was a feeding frenzy for invite codes, a testament to the hype OpenAI has been able to build in just the three years since it launched ChatGPT. Sora was launched with even greater ambitions — which is a problem for OpenAI, because demand for the app has slowed.

The logic behind Sora was fairly sound. Back in May, Meta launched an experimental AI content feed, which was, at the time, dubbed a “creepy” and “nightmarish” slop feed of AI garbage. But the philosophy behind it was pretty clear: AI content is meant to replace the user-generated content we see on our feeds every day.

This is already happening on platforms like YouTube, where we found earlier this year that four of the summer’s top 10 channels with the most subscribers consisted entirely of AI-generated videos, most of them being YouTube Shorts. TikTok-style short-form vertical videos seemed like a natural place for OpenAI to enter the content game. 

Other platforms, like Instagram, have bet big on overhauling their apps to function more like TikTok, so much so that Instagram Reels are now the primary type of posts all Instagram users see. But Sora wasn’t just meant to be OpenAI’s competitor to TikTok; it was designed to interface seamlessly with all these other TikTok-like feeds. 

Users could easily turn videos generated by Sora into Reels, Shorts, TikToks, Snaps, and more. Each repost would be free publicity for Sora, even if the views and engagement went to its rivals. OpenAI seemed to think it could generate the same AI content wave for video that it had unleashed for text and images. The Sora app itself doesn’t publicize any engagement totals, though, so those cross-platform reposts are the main tool we have for tracking the actual popularity of videos made on the app.

And those numbers were stratospheric — at least to begin with.

Accounts like @bestsoravids on Instagram and Epic Rankz on YouTube have recorded millions of views, likes, and comments on their Sora reposts. But all of their most successful posts happened in the week or two after the app’s launch, at the height of its visibility. In the first half of December, for example, only 20 videos with #sora or #sora2 in the hashtags received 1 million views or more on YouTube. Other popular Sora reposting accounts, like the previously mentioned @bestsoravids, stopped uploading videos entirely after a few weeks.

The other metric we have for measuring the health of Sora is downloads of the app itself. Those have fallen off. 

Remember, it took Sora just five days to get to 1 million total downloads, despite being invitation-only and iOS-only. Since that initial blitz, the iOS app was downloaded just over 5 million more times over the next roughly three months, according to data from Appfigures.

The iOS version launched on the last day of September, and it was downloaded nearly 2.7 million times through the end of October. In November, the monthly total was 1.9 million, and in December it was 1.5 million. (The Android version became available in early November. It was downloaded 1.4 million times in November and 1.7 million times in December.) 

For comparison, in December, TikTok was downloaded over 18 million times around the world, and YouTube was downloaded 5 million times despite having been an app for nearly two decades, literally available on iOS 1. It’s even more notable when you look at downloads in the US, specifically, which is by far Sora’s biggest market. 

After Sora was made available on Android on November 4, its daily download numbers were behind TikTok within a week. 

It’s possible, though, that Sora wasn’t a social media play at all, but rather a shrewd bit of advertising to entice a partner for something greater than the title of “the next TikTok.”

Walt Disney and Mickey Mouse, Vanity Fair 1933
Walt Disney had to draw his characters by hand, but you’ll just be able to use Sora (Edward Steichen/Getty Images)

OpenAI and Disney announced in December that they were partnering on a deal to let Sora users have exclusive access to Disney AI inside the video-generating app. But this, too, will backfire almost immediately unless serious safeguards are put in place. As Futurism has reported, users were already generating Pixar versions of 9/11, Jeffrey Epstein, and George Floyd’s death before the partnership was announced. They’re surely primed to find ways to misuse Disney IP once it’s added officially to the app.

This is the real pattern we continually see over and over with AI products like Sora: a massive wave of curiosity, a tremendous drop-off as normal people get bored, and a remaining user base of obsessed power users and trolls. Only this time, Disney has been duped into offering up its entire creative catalog to the denizens still left on the app.


Garbage Day is an award-winning newsletter that focuses on web culture and technology, covering a mix of memes, trends, and internet drama. We also run a program called Garbage Intelligence, a monthly report tracking the rise and fall of creators and accounts across every major platform on the web. We’ll be sharing some of our findings here on Sherwood News. You can subscribe to Garbage Day here.

More Tech

See all Tech
tech

Anthropic has surged past OpenAI in capturing business spending on generative-AI software

Last quarter, Anthropic attracted the lion’s share of trackable business spending on generative-AI software, according to new data from Ramp, a fintech company that provides corporate cards and expense management software for small firms and Fortune 500 companies alike.

The data showed that in the first quarter, Anthropic saw 37% of spending, its biggest share yet, versus 33% for OpenAI. Notably, the dataset doesn’t capture spending by Google or Microsoft.

OpenAI, which makes ChatGPT, still leads in overall adoption at 81% of AI buyers, but Anthropic is catching up, at nearly 63% in March. Overall, more than half of Ramp’s customers currently pay for AI, up from just 18% two years ago.

Anthropic’s enterprise tools, including Claude Code and Cowork, have been making waves among the business class, sending its revenue soaring.

Anthropic’s revenue share is even higher among companies spending on AI for the first time.

“Anthropic has definitely been on a tear,” Ara Kharazian, Ramp’s economist, told Sherwood News. “Its increase in adoption rates has been driven by its ability to sell to less technical users and smaller contracts than it typically has.”

It’s notable that midway through the first quarter, Anthropic had a falling-out with one of its biggest customers, the US government, which near the end of February decided to shun Anthropic’s products and lean into working with OpenAI.

tech

Report: Google ditches its objection to defense work, pitches Gemini to Pentagon

In 2018, Google employees protested against the company’s tech being used for the US military’s Project Maven — a drone targeting program — reminding the company of its “don’t be evil” motto.

After the controversy, the company declined to renew the contract with the Pentagon, drawing a bright line between Big Tech and the national security establishment.

What a difference a few years makes.

Google is now actively working to get its Gemini AI model to be used in classified national security settings, according to a new report from The Information. Seeking a similar deal to the one OpenAI hashed out with the Pentagon, Google reportedly wants a contract that allows use of Gemini in classified work, but with a prohibition on mass domestic surveillance and autonomous lethal weapons.

But Google is playing catch-up in a major way. Amazon and Microsoft both have been widely used for classified defense work, and contractors are already experienced in working with their cloud systems, while Google’s services have never been used in classified work.

What a difference a few years makes.

Google is now actively working to get its Gemini AI model to be used in classified national security settings, according to a new report from The Information. Seeking a similar deal to the one OpenAI hashed out with the Pentagon, Google reportedly wants a contract that allows use of Gemini in classified work, but with a prohibition on mass domestic surveillance and autonomous lethal weapons.

But Google is playing catch-up in a major way. Amazon and Microsoft both have been widely used for classified defense work, and contractors are already experienced in working with their cloud systems, while Google’s services have never been used in classified work.

1 in 5

We knew Tesla had been off-loading its struggling “apocalypse-proof” Cybertrucks onto CEO Elon Musk’s other companies, but now we know just how many.

The EV company sold about one in five Cybertrucks registered in the US in the fourth quarter to Musk’s other ventures, according to Bloomberg, citing data from S&P Global Mobility. The lion’s share went to SpaceX, which accounted for 1,279 of the 7,071 total registrations, while another 60 went to xAI (now part of SpaceX), Neuralink, and The Boring Company. All told, these inter-company sales represent roughly $100 million in value, and a vital lifeline for a vehicle that has failed to gain traction with the public, forcing Tesla to scale back production.

Musk’s companies have continued to scoop up the stainless steel behemoths this year, with another 158 Cybertruck purchases in January and 67 in February.

tech

TSMC CEO on Tesla and Intel’s Terafab: “There are no shortcuts”

Tesla CEO Elon Musk has reportedly asked the chip industry suppliers for his Terafab chipmaking project to move at “light speed” in an effort to help Tesla and SpaceX manufacture the AI chips they need.

On the company’s last earnings call, Musk said chip supply would be the “limiting factor” for Tesla’s growth in about three or four years. During a presentation for Terafab last month, Musk said, “We either build the Terafab or we don’t have the chips.” More established chipmaker Intel has since joined the effort.

Still, the world’s largest chipmaker isn’t convinced that “light speed” is physically possible. Speaking on an earnings call this morning, TSMC Chairman and CEO CC Wei offered a blunt assessment of Terafab’s ambitious timeline: “There are no shortcuts.” According to Wei, the physics of a modern foundry, which he says takes roughly five years to build and ramp, remains the ultimate speed limit, regardless of the customer’s urgency. “That’s a fundamental of the foundry industry,” he said.

Wei noted that Tesla remains a TSMC customer.

Latest Stories

Sherwood Media, LLC produces fresh and unique perspectives on topical financial news and is a fully owned subsidiary of Robinhood Markets, Inc., and any views expressed here do not necessarily reflect the views of any other Robinhood affiliate, including Robinhood Markets, Inc., Robinhood Financial LLC, Robinhood Securities, LLC, Robinhood Crypto, LLC, Robinhood Derivatives, LLC, or Robinhood Money, LLC. Futures and event contracts are offered through Robinhood Derivatives, LLC.