Sam Altman’s confident media run finally cracked on a recent Bloomberg Tech podcast. At one point in his recent interview on the BG2 podcast, the host asked Sam how a company with $13B in revenue can afford $1.4T in commitments. Within seconds, Altman replied “Happy to find a buyer for your shares.”

When pressed for clear numbers on how they would generate enough revenue, Altman cut the exchange short with a terse “Enough,” insisting OpenAI is focused on building “the infrastructure of intelligence.” Supporters saw conviction; skeptics saw evasion.

Beneath that bravado is a costly truth: OpenAI’s infrastructure deals, including a $38 billion, multi-year AWS partnership, commit billions long before profits arrive, fueling doubts about whether ambition alone can keep up with the bill.

Is OpenAI chasing a true scientific breakthrough or inflating a speculative bubble? Altman says automation will eventually generate returns by reinventing labor itself—AI coding AI, science done at scale, knowledge turned into product.

Critics counter that the company risks monetizing potential intelligence long before it appears.

Altman’s unease may not have been fear of failure, but recognition that investors are no longer content with poetic answers. For all his talk of alignment and safety, the biggest challenge may be financial grounding, not philosophical control.

As with any digital ad campaign, the important thing is to reach streaming audiences who will convert. Roku’s self-service Ads Manager stands ready with powerful segmentation and targeting — plus creative upscaling tools that transform existing assets into CTV-ready video ads. Bonus: we’re gifting you $5K in ad credits when you spend your first $5K on Roku Ads Manager. Just sign up and use code GET5K. Terms apply.

Use code GET5K now

“Fake news” used to mean a slanted headline. Now it can mean a newsroom getting fooled by a synthetic clip. Over the weekend, Fox News published a piece claiming SNAP beneficiaries threatened to ransack stores during a shutdown, then quietly rewrote it after observers showed several cited clips were AI generated.

An editor’s note now acknowledges the error. Newsmax ran a similar segment. The quick about-face shows how fast convincing fakes can jump from fringe feeds to national outlets before fact-checking catches up.

It is not only newsrooms. Meta’s chief AI scientist Yann LeCun posted a short video that appeared to show a New York City officer telling federal agents to “back off now.” 

Commenters quickly flagged telltale synthetic artifacts and suggested it was model made footage. The episode underlines a hard truth. People who know the tech can still be caught by short, highly polished clips that exploit our tendency to believe what fits a familiar narrative.

Modern diffusion and face swap pipelines now render micro-expressions, head turns, and voice cadence with enough fidelity to defeat quick human judgment and many automated filters. Platforms are experimenting with watermarking and content credentials, but enforcement is spotty. 

A recent test embedding provenance data in an AI video found that most social apps stripped or hid the labels, leaving viewers with no clear signal that they were watching a fake. Until labels are preserved end to end, detection will trail creation.

Treat viral video like a claim that needs evidence. Cross-check the clip across reputable outlets, run a reverse image or frame search, and scan for small inconsistencies in reflections, lighting, and lips that fall out of sync during fast phrases.

When a source corrects a story, update your priors instead of doubling down. Most important, be mindful of confirmation pull. The fakes that get us are the ones we want to be true.

AI is not turning malicious. It is turning efficient at producing persuasive forgeries that exploit human shortcuts and platform gaps. Until provenance survives uploads and detectors improve, assume any high impact clip is unverified unless proved otherwise.

🌍 China’s Xi seeks global AI cooporation at APEC

🧱 SLB’s new AI product takes aim at the energy industry

🚀 OpenAI & Amazon continue the AI spending spree with $38 billion deal

💻 AI companies prop up Wall Street as the rest falter

🎯 Europe wants more control of its AI future

🧨 Trump refuses Nvidia’s top chips for China and others

Here’s what to do:

🧠 Step 1: Pick a video clip you recently watched on social media—news, sports, or celebrity.

🔍 Step 2: Take a screenshot and run it through an AI image detection tool such as aiimagedetector.com.

🎨 Step 3: Ask an AI generator to recreate the same scene using a text prompt like “create a realistic news anchor talking about election results.”

🧩 Step 4: Compare both side by side and write down which details gave away the fake first—lighting, timing, or emotion.

By the end of this challenge, you’ll see how hard it’s getting to spot the synthetic world forming around us.

Click below ⬇️

Did Sam Altman take it too far, or is he really worried about the world uncovering OpenAI’s big secret? And, if Fox News and Newsmax can fall for deepfakes, what hope do we have? See you next time! 🚀

Stay informed, stay curious, and stay ahead with Jumble!

Zoe from Jumble