OpenAI’s Social Sora Gamble - DTNS 5117

Plus Andy Beach makes the connection between YouTube’s creator and AI strategies. [22:04]

Starring Tom Merritt, Jenn Cutter, and Andy Beach.

TOM: This is the Daily Tech News for Thursday, October 2, 2025. We tell you what you need to know, follow up on the context of those stories and help each other understand.

JENN: Today Is OpenAI’s Sora app a hellscape or a playground? Why not both?

I’m Tom Merritt,

I’m Jenn Cutter.

TOM: Let’s start with what you need to know with the big story.

[[BIG STORY]]
[[SOLO story of the day. Basic details, monitor commentary and sound when possible.]]

"OpenAI Completes Share Sale at Record $500 Billion Valuation - Bloomberg"
"OpenAI’s Sora 2 Copyright Infringement Machine Features Nazi SpongeBobs and Criminal Pikachus"
"OpenAI's new social app is filled with terrifying Sam Altman deepfakes | TechCrunch"
"OpenAI staff grapples with the company's social media push | TechCrunch"

TOM: OpenAI is in the news for being valuable and viral. First the value. It completed a valuation of the company, necessary to let employees cash in some of their shares since the company is not public. That put the company's value at $500 billion, making it the highest-valued non-public company, ahead of SpaceX. And that's up from the $300 billion it was valued at earlier this year when it conducted a round of financing.

OpenAI has a little viral buzz because of its new app Sora, which is generating more interest than Meta's similar feed in Meta AI. Sora, the app, lets you use Sora 2, the model, to generate 10-second videos that can include yourself. You scan your own image into make a "cameo". You can also add your friends or others if they allow you to.

404 Media's Jason Koebler has an article about all the copyrighted material that can be found in Sora, usually SpongeBob, Nintendo characters, and Rick and Morty. Some of those are made in tasteless ways. TechCrunch points out that there is an endless supply of fake videos of OpenAI CEO Sam Altman. Granted, every video on the app is fake. And Altman made his cameo available for anyone on the app to use.

However, in my personal experience, I found that even prompts about myself delivered “This content may violate our guardrails concerning third-party likeness.” As did attempts to make Luke Skywalker appear, as I had seen in someone else's post.

To their credit, OpenAI scientists John Hallman and Boaz Barak, both posted on X that they had concerns about the app. Barak posted, “Sora 2 is technically amazing but it’s premature to congratulate ourselves on avoiding the pitfalls of other social media apps and deepfakes.”

[[DISCUSS TOM’S EXPERIENCE ON SORA]]

"Google AI Overview appears to block results on searches for ‘Trump cognitive decline’ but not for Biden"

TAG OUT
TOM: Meanwhile, Google is also in hot water with the Verge, discovering that a search for “is trump in cognitive decline,” does not deliver an AI overview and only returns a list of links in AI mode. Whereas a similar search for “is biden in cognitive decline,” will return an overview and an answer in AI mode. Google told The Independent, “We don't show AI Overviews on every query and similarly in AI Mode, for some topics (like current events) we may show a list of links as the response.” My personal theory is that if the ratio of slop articles is too high, it filters out responses. There is a longer history of reliable reporting on President Biden's mental health and quite a lot of speculation on President Trump's. In my own testing, most political leaders like Republican Senator Mitch McConnell, do show an AI overview. However, a search including the name of Democratic Senator John Fetterman did not show an overview or a response in AI mode.

JENN: DTNS is made possible by you the listener. Thanks to
thatCharlieDude
Justin Zellers
Carmine Bailey
New Patrons: Onewhoknows and Martin
Returning Patron: Kelly (annual)
[[BREAK]]
[[PAUSE]]

TOM: There’s more we need to know today, let’s get to the briefs.

[[BRIEFS]]
[[3-9 more solo reads with sound to complete the day in tech news. These are informational with minor commentary.]]

"Apple Shelves Vision Headset Revamp to Prioritize Meta-Like AI Smart Glasses - Bloomberg"
"FCC Accidentally Leaks Apple's Next Vision Pro - MacRumors"
"M5-powered iPad Pro breaks cover in GeekBench, scoring 4,133 in single-threaded tests — matches M4 Max and beats every single-core PC chip score | Tom's Hardware"

JENN: 9to5 Mac reports that on Tuesday, the FCC made public documents that refer to an Apple-designed "Head Mounted Device" with an illustration that looks like an Apple Vision Pro. However, on Wednesday, Bloomberg's Mark Gurman reported that Apple has moved people from working on a smaller, lighter version of the Apple Vision Pro, to work on smart glasses to compete with Meta Ray-ban's and the dozens of other competitors about to enter that market. Gurman's sources say there are two models. The N50, which pairs with an iPhone and does not have a display, is due to come out next year, and another model, with a screen, is planned for 2028. Gurman says that the updated Vision Pro with a new chip, likely the one in the FCC filings with an M5 chip, is expected by the end of the year. And, speaking of the M5 chip, Tom's Hardware notes that an iPad with an M5 has shown up on GeekBench with a score that matches the M4 Max and beats all other PC single-thread scores.

"Exclusive: Mira Murati’s Stealth AI Lab Launches Its First Product | WIRED"

TOM: Former OpenAI CTO Mira Murati's new company, Thinking Machine Labs, has announced its first product. Tinker can automatically create custom Frontier models for companies looking to narrow general-purpose models to their specific needs. Organizations spend a lot of effort now to customize models for tasks like drafting legal agreements reliably, answering medical questions, or solving math or accountancy problems. Tinker lets you fine-tune Meta's Llama or Alibaba's Qwen by writing a few lines of code with labeled data or reinforcement learning. You can then download the refined model to run on your own system.

"Japan's largest brewer suspends operations due to cyberattack"

JENN: Thanks to motang for tipping us off to this story on the DTNS subreddit. Japan's Asahi Group Holdings disclosed that a malicious attack on its systems has disrupted ordering, shipping, and customer service activity. Asahi is one of the world's biggest breweries and holds one-third of the market for beer in Japan. Asahi says that no personal information was accessed, and it is working to restore systems, but did not give a timeline. Meanwhile, convenience stores eye their dwindling stock with concern.

"Instagram head says company is not using your microphone to listen to you (with AI data, it won't need to) | TechCrunch"
"Adam Mosseri’s ‘we’re totally not spying on you’ video is raising a lot of questions | The Verge"

TOM: Instagram head Adam Mosseri posted on Wednesday to combat the idea that Instagram listens to your conversations to serve you ads. He said, "We do not use the phone's microphone to eavesdrop on you." Mosseri offered three possibilities for the coincidences that make people think they do. 1. You interacted with a topic elsewhere at a place that shares information with Instagram, 2. You were talking to a friend who had interacted with the product, and Instagram knows your friends and served you the ad or 3. You had seen the ad before and not realized it. Facebook has denied the idea before in a blog post in 2016, at a US Senate hearing in 2018, and in a support document on its website.

"Spacecoin beams blockchain transaction through space in bid for decentralized internet | TechCrunch"

JENN: Spacecoin says it has successfully tested transmitting a secure blockchain record through its Endurosat CTC-0 satellite in orbit, between Chile and Portugal. Founder Tae Oh said the test was part of his hope to "build a decentralized network of communications satellites and fulfill the vision of building permissionless connectivity.”

"Meet the Arc spacecraft: It aims to deliver cargo anywhere in the world in an hour - Ars Technica"

TOM: A company called Inversion announced its "on demand" delivery vehicles at a factory in Los Angeles on Wednesday. These are not sidewalk robots or autonomous cars, their spacecraft. Inversion's Arc spacecraft can hold up to 225 kilograms of supplies in space for up to 5 years. When needed, they can descend and deliver them in less than an hour, anywhere on Earth. So you're not going to have burritos delivered by Arc. The first client is the US military, which could use it for medical supplies, drones, or other cargo. Inversion plans to launch the first Arc vehicle into space by the end of next year.

"Amazon’s Vega OS launch trick: cloud-streamed apps | The Verge"

JENN: The Verge reports that Amazon's new Vega OS, used on its new FireTV stick, will have all the apps you expect from a device, even though it no longer uses Android as the OS. That's because apparently it will stream the entire app from the cloud, not just the content.

TOM: And finally, some quick headlines that are just good to know if you want to understand the news in the future.

"Exclusive-Yahoo nears deal to sell AOL to Italy's Bending Spoons for $1.4 billion, sources say"

JENN: Reuters sources say that Bending Spoons, home of brands like Evernote, Streamyard and Meetup, is in talks to buy what's left of AOL from Yahoo.

"Spotify now lets you exclude tracks from your taste profile, improving recommendations | TechCrunch"

TOM: Spotify announced a new feature on Wednesday that lets you exclude certain music tracks from the recommendation algorithm, so you can safely let your niece play the Pokémon theme song 100 times in a row, and then remove it later.

"Music labels close to landmark AI licensing deals"

JENN: The Financial Times reports that Universal Music and Warner Music are close to multiple deals that would pay them royalties in exchange for the use of their music in AI models from companies like Google, ElevenLabs, Stability AI, Suno, and Udio.

"Brave Now Has Over 100 Million Users - Thurrott.com"

TOM: The Brave browser, which emphasizes privacy, says it now has more than 100 million active users across mobile and desktop, about half what Firefox has.

"Perplexity’s Comet AI browser now free; Max users get new 'background assistant' | TechCrunch"

JENN: Perplexity is making its Comet browser, which builds in its LLM tools, available for free for all users, with rate limits.

"Walmart steps up automation with labour-saving sensors"

TOM: The Financial Times reports that Walmart will use Bluetooth sensors from Wiliot, to track 90 million pallets it uses to ship groceries to its stores.

"Hulu To Become International Tile On Disney+ On October 8"

JENN: Disney will replace its Star branding on its international streaming services with the Hulu brand, starting October 8th.

"Intel in early talks to add AMD as foundry customer | Semafor"

TOM: Semafor says Intel is in early talks to sign up AMD to its foundry business.

"Dutch court orders Meta to let users disable algorithmic timelines | NL Times"

JENN: A court in the Netherlands has ordered Meta to offer Facebook and Instagram users the option for a non-algorithmic timeline.

"T-Mobile’s satellite service now supports Google Maps, WhatsApp, and X | The Verge"

TOM: T-Mobile now supports using WhatsApp, Google Maps, X, and AccuWeather on its satellite-to-cell service.

"Typepad - Closed for Business"

JENN: The blogging service Typead closed up its business September 30th.

"Two Amazon Prime Air drones crashed in Arizona"

TOM: And on Wednesday morning, two Amazon Prime Air delivery drones collided with a construction crane in suburban Phoenix, Arizona.

Those are the essentials for today. Let’s dive a little deeper.

[[SEGMENT A - FROM SCHEDULE]]

JENN: Previously, we talked about YouTube making two big moves: doubling down on livestreaming and leaning into AI for creation. Those two efforts are connected. Andy Beach connects the dots.

[[PROMO]]

TOM: If you have feedback about anything that gets brought up on the show… Get in touch with us on the socials. @DTNSshow on X, Instagram, Threads!, Blue Sky, and Mastodon. For TikTok and YouTube, you can find us at Daily Tech News Show.

[[BREAK]]
[[PAUSE]]

[[HELPING EACH OTHER UNDERSTAND]]
[[Short missives from people with experience. Could be written email or pre-recorded from the person.]]

JENN: We end every episode of DTNS with some shared wisdom. Today Andrew has some observations about AI actors.

TOM:
Andrew writes:
In regards to Gen A.I. talent. I have begun to see it being used frequently here in Brazil. It's used for a lot of billboard ads and Social Media ads by small to medium size companies.

I modeled from my teens to my late 30's, and I did a lot of extra work (Movies) and print modeling (Magazines).

The commercials that I am seeing AI used in are replacing extras or actors who would have non speaking roles,

It seems like it's Veo3 or possibly Heygen. It doesn't look like a real person to me, but I am always looking for it.

I think any big company trying to use it would draw controversy.

[[DISCUSS]]

JENN: What are you thinking about? Got some insight into a story? Share it with us feedback@dailytechnewsshow.com

TOM: Thanks to Andy Beach and Andre for contributing to today’s show. And thank YOU for being along for Daily Tech News Show. You can keep us in business by becoming a patron, at Patreon.com/dtns