
Dear Readers,
The past 48 hours in AI felt like standing in the middle of a fast-rising tide, every time you blink, another wave hits. What do you do when Anthropic locks in a full 1GW compute deal, Nvidia and Microsoft pour $15B more into frontier models, and Google researchers casually say there’s “no wall in sight”? You take a breath, lean in, and start asking what this acceleration really means for the tools we’ll use, the power structures being built, and the future we’re all about to live in. Today’s issue follows that line of tension: between scale and capability, between hype and reality, and between what AI is now and what it’s rapidly becoming.
Inside, you’ll find a packed lineup: Anthropic’s historic partnership that reshapes the compute landscape, the valuation jump that signals a new race dynamic, Google’s internal confidence about Gemini 3’s leap, and the latest model upgrades, including Grok 4.1’s surprising emotional intelligence and the newest signals from OpenAI’s Codex-Max. We also break down the first big analyses of Gemini 3.0 Pro, highlight the most revealing chart of the day, and bring you a curated mix of prompts, videos, and rumors shaping the week. Dive in: today’s stories connect the dots in a way that will make you see the next chapter of AI a little more clearly.
In Today’s Issue:
🤝 Grok 4.1 delivers enhanced emotional intelligence and creativity
🚀 Prof. Ethan Mollick traces AI's rapid evolution
✨ Andrej Karpathy praises Gemini 3 as a "tier 1" model
🥇 Gemini 3 Deep Think is revealed as the powerful engine
✨ And more AI goodness…
All the best,




Anthropic Secures 1GW Compute Power
Anthropic has just announced a monumental strategic partnership with Microsoft and NVIDIA, locking in access to over 1 gigawatt of compute power and committing a massive $30 billion to scale their AI infrastructure, making Claude the only frontier model available across all major cloud providers (Azure, AWS, GCP).

Anthropic Valuation Doubles To $350B
Anthropic just secured a massive $15 billion investment from Microsoft ($5B) and Nvidia ($10B), rocketing its valuation from $183 billion in September to an incredible $350 billion, nearly doubling in just two months.

The delta between 2.5 and 3.0 is as big as we've ever seen. No walls in sight!
Gemini CO-LEAD and VP of Google Research confirms that the leap from 2.5 Pro pretraining to 3 Pro is as big as they’ve ever seen, while post-training is as rich as ever. There is literally no wall in sight, only acceleration.


A new era with Gemini 3 - a summary.


OpenAI is planning to introduce GPT-5.1-Codex-Max for bigger projects!
"Codex is now powered by GPT-5.1-Codex-Max, our latest frontier agentic coding model. It is smarter and faster than its predecessors and capable of long-running project-scale work."


Grok 4.1 released: A big upgrade!
The Takeaway
👉 Grok 4.1 offers major improvements in handling emotional context, stronger empathy, and a more conversational tone.
👉 It demonstrates significantly reduced hallucination rates, improving trustworthiness in information-seeking use cases.
👉 Widely available now, making it easy to test real-world applications in chatbots, assistive tools, or creative writing frameworks.
👉 Signals a shift: AI development now values social fluency alongside logic and accuracy.
The new model Grok 4.1 by xAI has just landed and it’s not just smarter, it’s friendlier, too. The update brings noticeably better emotional intelligence, creative writing skills, and drastically fewer hallucinations compared to its predecessor.

In practice, this means Grok 4.1 can pick up on subtle cues in your prompts (“I miss my cat so much it hurts”), respond with real empathy, and create richer, more human-feeling dialogue. For the AI-community, this matters because it blurs the line between purely functional AI assistants and ones that can engage in deeper, more nuanced human-centric conversation.

What’s really exciting is how this shifts the design space: builders and researchers can now aim for agents that are both technically competent and socially fluent.
Why it matters: It marks a major step toward AI that blends advanced reasoning with emotional awareness. For creators and researchers, it raises the bar for what a “useful AI companion” can look like.
Sources:


Make Every Platform Work for Your Ads
Marketers waste millions on bad creatives.
You don’t have to.
Neurons AI predicts effectiveness in seconds.
Not days. Not weeks.
Test for recall, attention, impact, and more; before a dollar gets spent.
Brands like Google, Facebook, and Coca-Cola already trust it. Neurons clients saw results like +73% CTR, 2x CVR, and +20% brand awareness.



Three Years from GPT-3 to Gemini 3
Prof. Ethan Mollick reviews the rapid three-year progression from GPT-3 to Google’s newly released Gemini 3, highlighting a fundamental shift from simple chatbots to autonomous “digital coworkers.” By assigning the model open-ended research tasks, he demonstrates that Gemini 3 can independently clean data, formulate hypotheses, and write sophisticated academic papers with minimal human guidance. This evolution turns the user’s role from a corrector of errors to a manager of agents, responsible for directing workflows and approving high-level decisions rather than micromanaging execution. Ultimately, Mollick concludes that we have entered a new era where AI agents can plan and build entire projects, requiring humans to provide strategic supervision rather than basic instruction.

Andrej Karpathy considers Gemini 3.0 to be a Tier 1 model.
Skeptical of public benchmarks due to the high incentive for developers to game them, Andrej Karpathy praises Gemini 3 as a “tier 1” model after finding it highly capable in writing, coding, and humor during personal testing. He advises users to judge models through direct interaction and to look forward to independent private evaluations to confirm these positive early impressions.

Gemini 3 Deep Think is the model behind Gold-Medal IMO and ICPC
“Gemini 3 Deep Think is next level. Deep Think was the engine behind our gold-medal-level wins at IMO and ICPC, and now powers an even stronger version of Gemini 3. SOTA on top of SOTA. More to come soon!”






