Apple's nightmare week

It’s Friday.

Apple plans to launch AI-powered smart glasses by late 2026, joining a growing race led by Meta and Google as its Vision Pro lags and broader AI strategy struggles to gain traction, Bloomberg reports.

OpenAI buys Jony Ive's startup for $6.5 billion

OpenAI just acquired io, the secretive AI device startup co-founded by Apple's legendary designer Jony Ive, for nearly $6.5 billion in stock. This is OpenAI's largest acquisition ever and signals their serious push into hardware.

The power couple: Sam Altman and Jony Ive are now working together to create what they call a "totally new kind of thing." Their first device is expected to debut in 2026. Ive, who designed the iPhone, iPad, and Apple Watch alongside Steve Jobs, called Altman a "rare visionary."

What OpenAI gets:

• About 55 hardware engineers, software developers, and manufacturing experts

• Ive and other former Apple designers behind iconic products

• A dedicated hardware unit led by Apple veterans including Evans Hankey (Ive's former successor) and Tang Tan (who led iPhone and Apple Watch design)

The bigger picture: This is a vertical integration play. OpenAI wants to control the entire AI stack from model to form factor to user experience. Less reliance on Apple, Google, and Microsoft means total control and maximum optimization.

Ive's LoveFrom design firm will also handle all of OpenAI's design work, including software interfaces. Altman admitted they're still in the "terminal phase of AI interactions" and haven't figured out what the graphical user interface equivalent will be for AI.

Apple's problem gets worse: Apple shares dropped 2.3% on the news. The company is already struggling in AI, relying partly on OpenAI's ChatGPT to fill gaps in their own AI platform. Now their former star designer is building competing products with one of their key AI partners.

The irony? Ive was once called Jobs' "spiritual partner." Now he's creating rival technology that could threaten the very devices he helped design. Altman believes Jobs would be "damn proud" of Ive's latest move.

What we don't know: Neither Ive nor Altman would reveal specifics about their hardware plans. The AI device market has seen notable failures like the Humane Ai Pin and Rabbit r1. Ive called those "very poor products" and noted there's been "an absence of new ways of thinking expressed in products."

The manufacturing wildcard: Trump just threatened a 25% tariff on iPhones not made in America, telling Tim Cook he expects iPhones "sold in the United States will be manufactured and built in the United States, not India, or anyplace else." This creates an interesting advantage for OpenAI - they can design their manufacturing strategy from scratch while Apple struggles with existing overseas production.

With Ive's design expertise and OpenAI's AI leadership, they're betting they can create something that works where others have failed. Hardware is becoming the new AI frontier, and OpenAI wants to be the Apple of artificial intelligence.

PRESENTED BY HEAD

50K WAITLIST IN 6 HRS! Head's Beta: Silicon Valley's Hottest Ticket!

The secret's out! An unprecedented 50,000 users stormed Head’s exclusive beta waiting list in just SIX hours. This explosive demand for the world's first AI Marketer has made beta invite codes the most coveted digital asset in Silicon Valley. Scarcity is real, and the frenzy is just beginning. Don't miss out on the revolution.

Limited Spots: See If You're In!

Google's AI creates videos so real they're scaring people

Google's new Veo 3 AI creates videos so realistic most people can't tell they're fake. Unlike OpenAI's Sora, Veo 3 includes dialogue, soundtracks, and sound effects while following complex prompts with frightening accuracy.

What makes it different: The AI follows real-world physics, offers accurate lip-syncing, and generates people with lifelike features (including five fingers per hand). The telltale signs of synthetic content are mostly gone.

The proof: Filmmaker Hashem Al-Ghaili posted AI-generated actors railing against their creators. The clips went viral because viewers couldn't tell they were fake. Available now for $249/month to Google AI Ultra subscribers.

The mystery: Google won't say how Veo 3 was trained, but when multiple users prompted stand-up comedy videos, it generated the same lame dad joke. This suggests potential training on copyrighted content, similar to how Sora appeared trained on specific YouTuber videos.

The world hasn't figured out how to handle authorship, consent, and rights when hyperrealistic fake videos become this easy to produce.

Anthropic claims the AI coding crown

Anthropic just launched Claude 4 Opus, which it claims is the world's best coding AI. The company says Opus can perform thousands of steps over hours without losing focus, while the new Claude Sonnet 4 improves on everyday coding tasks.

What's different: Both are hybrid models that can either answer immediately or take extra compute time for complex reasoning. Anthropic had to add new safety controls because "the model is so strong" according to their responsible scaling policy.

The timing: This comes after a busy week that included Microsoft's new coding agent, Google's AI search expansion, and OpenAI's $6.5 billion hardware bet. The race for "best frontier model" is heating up as questions persist about whether current AI techniques can reach new heights.

Anthropic is also changing how its models show their reasoning - now providing summaries of thought processes rather than documenting each step to "preserve visibility while better securing our models."