Gemini gets personal, Wikipedia AI licensing & more (January 15, 2026)

Good morning,

Big platforms are racing to personalize AI, while infrastructure, data rights, and open models are becoming the real battlegrounds. Today’s stories show how control over compute, content, and context is reshaping who actually wins in AI.

Let’s dive in 👇

━━━

Powered by Aloa • AI Research & Education Hub

━━━

Want to engage in conversation about AI news? Join the Byte-Sized Reddit Community

🤖 Platform Moves & Power Plays

🔍 Gemini gets personal

Google is rolling out “Personal Intelligence” inside Gemini, letting the model tap into emails, photos, and activity to tailor responses. The shift moves Gemini from a generic assistant toward a memory-driven system that understands users over time. It also raises familiar questions around data boundaries, consent, and how much context users are willing to trade for convenience.

📚 Wikipedia formalizes AI licensing, big labs already in

Wikipedia has unveiled a formal AI licensing program as part of its 25th anniversary, standardizing how AI companies use its content for training. The Wikimedia Foundation says Perplexity, OpenAI, Microsoft, Meta, and Mistral are already participating through approved, paid access channels. The model is designed to fund Wikipedia long-term while keeping the site free and editorially independent.

⚡ OpenAI locks in massive Cerebras compute for $10B

OpenAI reportedly signed a deal worth $10B for compute from Cerebras, signaling continued demand for specialized AI hardware. The move suggests OpenAI is diversifying beyond Nvidia to secure long-term capacity. It also reinforces how access to compute, not models alone, is becoming a core competitive moat.

🧪 Models, Money & Talent Shifts

💰 BlackRock and Microsoft deepen AI ties, raise $12.5B

BlackRock and Microsoft’s AI collaboration is reportedly targeting $12.5B in investment capacity focused on AI infrastructure. The partnership blends capital, cloud, and enterprise distribution at massive scale. It’s another sign that AI’s next phase is capital-intensive and increasingly financialized.

🧠 Thinking Machines Lab loses co-founders

Mira Murati’s startup, Thinking Machines Lab, is reportedly losing two co-founders back to OpenAI. The moves underline how fluid top AI talent remains between frontier labs and new startups. It also shows how hard it is for early teams to compete with OpenAI’s resources and gravitational pull.

🖼️ Open-source image model beats Google

Z-AI’s open-source GLM-Image model reportedly outperforms Google’s Nano Banana Pro on complex text rendering. The result reinforces that open models can compete, and sometimes win, on quality. It adds pressure on closed providers to justify proprietary advantages beyond branding.

🛠️ Tools of the Day

→ Teamout – Shared team calendar built for distributed and remote teams.
→ Folk – Lightweight CRM designed around relationships, not pipelines.
→ LyzrGPT – Framework for building custom, domain-specific AI agents.

⚡ Quick Hits

→ TSMC Q4 profit hits record as AI chip demand surges.
→ Skild AI reaches $14B valuation in robotics push.
→ Apple and Google drop Grok from app stores after backlash.
→ ChatGPT Translate launches as a standalone web tool.
→ Volvo backs Gemini for next-gen in-car AI.
→ Philippines joins bans on Musk’s Grok AI.
→ Gemini expands memory across photos and email.
→ India’s Emversity doubles valuation on human-first skills.
→ Bandcamp bans AI music entirely from the platform.
→ AI data centers stall due to skilled labor shortages.

🧾 TL;DR

AI competition is shifting from model releases to control over data, compute, and personalization. Google is betting on deeper user context, while Wikipedia and governments push back on unchecked AI expansion. Meanwhile, capital, talent, and open-source breakthroughs continue to reshape where real leverage in AI lives.

Cheers,
David