Meta's chief AI scientist exits with sharp criticism

PLUS: OpenAI locks 40% global memory, Grok faces EU backlash. Alibaba rivals NanoBanana Pro & Plaud ships $179 NotePin S AI wearable.

In today’s agenda:

1️⃣ Yann LeCun departs Meta after decade, calls replacement "inexperienced" and reveals Llama 4 benchmarks "fudged"

2️⃣ OpenAI secures 40% of global DRAM supply, DDR5 prices surge 300% to $710

3️⃣ Grok faces action from France, UK, India, Malaysia over AI "undressing" features including minors

  • Alibaba's Qwen-Image-2512 ranks as strongest open-source generator, rivaling Google’s NanoBana Pro

  • Plaud unveils $179 NotePin S wearable and desktop app to capture both in-person and remote meetings

MAIN AI UPDATES / 5th January 2026

🚪 Meta's chief AI scientist exits with sharp criticism 🚪
LeCun blasts new leadership on his way out

Yann LeCun, Meta's chief AI scientist for over a decade, departed the company with pointed criticism in a candid FT interview. He called his replacement Alexandr Wang "inexperienced" and lacking research credentials, admitted that Llama 4 benchmarks were "fudged a little bit," and predicted more departures from Meta's GenAI team. LeCun revealed he'll serve as executive chair of AMI, his new AI venture, while maintaining that LLMs are a "dead end" for achieving superintelligence—directly contradicting Meta's current strategy.

🔌 OpenAI locks 40% global memory supply 🔌
Tech's hidden tax on electronics

In October, OpenAI secured preliminary agreements with Samsung and SK Hynix for 900,000 DRAM wafers monthly—roughly 40% of global output—to power its Stargate data center initiative. The massive deal has sparked a memory shortage, with DDR5 memory prices surging nearly 300% from $180 to $710. Experts expect prices to keep rising through 2026, affecting everything from smartphones to PCs as the AI infrastructure race creates ripple effects across consumer electronics.

⚠️ Grok faces EU backlash over 'undressing' AI ⚠️
Multiple European governments condemn xAI's image editing capabilities

xAI's Grok is facing government action from France, India, Malaysia, and the UK after users flooded X with requests to digitally undress people—including minors—using the model's AI editing capabilities. France called the outputs "clearly illegal" under the EU's Digital Services Act, while Musk stated users creating illegal content "will suffer the same consequences." The controversy highlights the challenges of unrestricted AI editing powers at scale, with X's anonymous accounts and global userbase making enforcement extremely difficult.

INTERESTING TO KNOW

🖼️ Alibaba rivals Gemini with new image AI 🖼️

Alibaba's Qwen team released Qwen-Image-2512, addressing enterprise pain points: realistic human rendering without the "AI look," accurate text embedding for slides, and finer natural textures. The model ranked as the strongest open-source image generator, rivaling Google Gemini 3 Pro in head-to-head comparisons. The release signals China's continued push to match Western AI capabilities in multimodal generation.

🎙️ Plaud ships $179 NotePin S AI wearable 🎙️

Ahead of CES, Plaud launched the NotePin S, a $179 AI-powered wearable notetaker featuring a physical recording button, clip and lanyard accessories, and Apple Find My integration. Alongside the hardware, Plaud introduced a desktop app for digital meeting transcription that captures system audio and applies AI structuring. The rollout positions Plaud as a direct competitor to established meeting AI tools like Granola and Fireflies, which matters for users seeking alternatives to always-on recording approaches. The physical button design prioritizes explicit user control over passive capture.

📩 Have questions or feedback? Just reply to this email , we’d love to hear from you!

🔗 Stay connected: