In partnership with
shared.image.missing_image
shared.image.missing_image
Good morning, {{first_name | AI enthusiast}}.
Meta used its Connect keynote to show what wearable AI looks like once you take the camera away: a new pair of Ray-Ban glasses built entirely around voice, launched right alongside the company's first-ever VR glasses.
It's a bet that people will wear AI on their face all day only if it doesn't double as a surveillance device — and Meta backed that bet with five hardware announcements at once. Will ditching the camera actually win over skeptical buyers, or does AI on your face still need something to see?
Today in AI Brief:
Meta's camera-free glasses and first VR headset
Alibaba's new chip and trillion-parameter ambitions
Mirage's free video editor built for AI agents
One idea shouldn't take six rewrites to post.
Posting everywhere means rewriting one idea six times, so you post to one, or none. SureThing turns one idea into native posts for every platform.
Meta Unveils Camera-Free Glasses and Its First VR Headset
In Brief: Meta unveiled a wave of new hardware at Connect 2026, headlined by Ray-Ban Meta Audio — the company's first camera-free smart glasses — alongside Meta VR Glasses, its first standalone VR headset built for everyday wear.
The Details:
Ray-Ban Meta Audio ships October 13 for $349, weighs 43 grams, and runs up to 12 hours on a charge, trading the camera for an all-day voice-and-AI design in Clubmaster and Burbank styles.
Meta VR Glasses cost $1,299.99 and launch spring 2027, weighing about 100 grams — five times lighter than the Quest 3 — with a 5K micro-OLED display and a detachable battery-and-processor puck.
The refreshed Ray-Ban Meta Gen 3 adds a six-microphone array that cuts background noise over 90%, and a new Muse Charm pendant teases a pocket-sized AI companion coming later this year.
Take Away:
Meta is betting that removing the camera — not adding more AI — is what finally makes smart glasses an everyday habit, especially as regulators and privacy advocates scrutinize covert recording. If the audio-only glasses catch on where camera versions haven't, it reshapes what "AI wearable" means for every other company chasing the same form factor.
Alibaba Unveils Its Most Powerful AI Chip Yet
In Brief: Alibaba unveiled the Zhenwu V900, a new AI chip CEO Eddie Wu called the most powerful in China, alongside plans to train a future model with 5 trillion to 10 trillion parameters.
The Details:
The Zhenwu V900 delivers 3x the performance of its predecessor and supports clusters of up to 500,000 chips training and running models together, with mass production targeted for early 2027.
Alibaba's flagship Qwen3.8-Max currently runs on 2.4 trillion parameters; the planned Qwen 4.5 and Qwen 5 models would scale that up to roughly 4x larger, closing ground on the biggest US models.
The announcement lands just ahead of Xi Jinping's state visit to Washington, where Alibaba's push to build domestic chips despite ongoing US export limits is expected to be a talking point.
Take Away:
China's chipmakers keep closing the gap Washington's export controls were designed to protect, and a 10-trillion-parameter model would put Alibaba within reach of the largest models US labs have shipped. Expect chip and model announcements like this to keep landing on a predictable cadence as the AI race becomes as much about semiconductor independence as model quality.
SPONSORED BY THE CODE
100+ Claude Code hacks to ship code 10X faster
Top engineers at Anthropic and OpenAI say AI now writes 100% of their code.
If you're not using AI, you're spending 40 hours doing what they do in 4.
These 100+ Claude Code hacks fix that and help you ship 10x faster.
Sign up for The Code and get:
100+ Claude Code hacks used by top engineers — free
The Code newsletter — learn the latest AI tools, tips, and skills to code faster with AI in 5 minutes a day
Mirage Launches a Free Video Editor Built for AI Agents
In Brief: Mirage launched Tesseract, a free video creative suite that gives AI agents direct access to professional editing tools — layers, keyframes, compositing, and sound — instead of forcing them to fake edits through code.
The Details:
The engine runs locally on Mac, Windows, and Linux, letting agents like GPT-6 Astra and Sol trim clips, build motion graphics, and mix audio without opening Premiere or After Effects.
Everything is free, including rendering — users only pay for whichever AI agent subscription they're already using to direct the edits.
Mirage is backed by General Catalyst, Andreessen Horowitz, and Sequoia Capital, and positions Tesseract as a way to skip paid editing subscriptions entirely when an agent is doing the cutting.
Take Away:
Video editing has been one of the last creative tasks agents couldn't touch without a human at the controls, and Tesseract closes that gap by handing them the same tools professional editors use. If agents can now finish a cut end-to-end, expect video production costs to keep falling for anyone willing to describe the edit instead of making it by hand.
Everything else in AI
Firecrawl raised $75M in a Series B and launched Alexandria, a unified data platform giving AI agents access to 113 million technical documents across 82 data providers.
Cyera secured another $400M from Goldman Sachs at a $12B+ valuation, building tools that track what AI agents can actually reach inside enterprise data.
Andreessen Horowitz launched a tuition-free "Academy" for high school grads with no tests or grades, betting $42M that project-based AI training beats a traditional college degree.