Good morning, AI enthusiast.

Google DeepMind's new AI has moved beyond analysis to generate a scientific breakthrough of its own. The model proposed a new way to make hard-to-treat cancers visible to the immune system, a finding that has since been validated by university researchers in a lab.

This marks a significant shift from AI simply analyzing data to actively proposing testable ideas for research. What could this new paradigm mean for accelerating discovery across other complex scientific problems?

In today’s AI recap:

• Google DeepMind's AI-generated cancer therapy

• Claude’s new reusable ‘Skills’ feature

• TSMC’s record profits from AI chips

• The publisher revolt against AI Overviews

• 8 trending AI Tools

Google's Cancer Clue

The Recap: Google DeepMind's new AI generated a novel hypothesis for making hard-to-treat cancers visible to the immune system, a finding since validated by Yale University researchers in a lab. The breakthrough could unlock new combination therapies for previously untreatable tumors.

Unpacked:

• The AI pinpointed how to turn "cold" tumors "hot" by predicting that the drug silmitasertib, combined with low-dose interferon, could boost a tumor's immune signals by nearly 50%.

• This discovery comes from the massive C2S-Scale 27B model, a system built on Google's Gemma and trained to understand the "language" of individual cells.

• This marks a significant shift from AI simply analyzing data to actively generating testable ideas, suggesting a future where AI accelerates scientific discovery by proposing its own experiments.

Bottom line: This achievement showcases AI's potential to act as a creative partner in scientific research, not just an analytical tool. It offers a blueprint for how AI can help solve some of biology's most complex problems much faster than traditional methods allow.

Go from AI overwhelmed to AI savvy professional

AI keeps coming up at work, but you still don't get it?

That's exactly why 1M+ professionals working at Google, Meta, and OpenAI read Superhuman AI daily.

Here's what you get:

• Daily AI news that matters for your career - Filtered from 1000s of sources so you know what affects your industry.

• Step-by-step tutorials you can use immediately - Real prompts and workflows that solve actual business problems.

• New AI tools tested and reviewed - We try everything to deliver tools that drive real results.

• All in just 3 minutes a day

Join 1M+ pros

AI Tools of the Day

1. 🤖 SmythOS – Build and deploy powerful autonomous agents across platforms like ChatGPT, Slack, and Discord using a drag-and-drop visual interface.

2. 🏄 Surf.new – Unleash autonomous web agents that can research, compare prices, and interact with websites just like a human.

3. 📊 TEXT2SQL.AI – Instantly translate your plain English questions into complex SQL queries, Excel formulas, and Regex expressions.

4. 🎵 MizanMe – Generate personalized, science-based motivational music that adapts to your unique mood and goals for peak focus and performance.

5. 🎬 Jupitrr AI – Automatically transform your simple talking head videos into engaging, shareable content by intelligently adding B-roll and captions.

6. 🗣️ Vozo – Effortlessly dub your videos into any language with flawless AI-powered lip-syncing for a truly global audience.

7. 🛡️ WitnessAI – Secure your company's AI adoption by monitoring LLM usage, enforcing policies, and preventing sensitive data leaks in real-time.

8. 💻 Swimm.io – Eliminate technical debt by using an AI-powered assistant that understands your codebase and automatically generates up-to-date documentation.

Explore the Best AI Tools Directory to find tools that will 10x your output 📈

Claude's New Superpower

The Recap: Anthropic just launched Skills, a new feature that lets you package instructions, resources, and code into reusable modules for Claude. This allows you to teach the AI specialized workflows and knowledge to handle complex tasks with greater precision.

Unpacked:

• At their core, Skills are simple folders containing a central SKILL.md file, making it easy for anyone to bundle and share procedural knowledge without building a custom agent from scratch.

• The system uses progressive disclosure to intelligently load only the necessary information into the context window, preventing bloat and keeping interactions highly efficient.

• These reusable capabilities work across the entire Claude ecosystem—including the web app, API, and Claude Code—and can even contain executable Python scripts for deterministic tasks.

Bottom line: This update moves Claude beyond a generalist tool, enabling users to build a library of specialized experts tailored to specific jobs. It also paves the way for a future marketplace where users could share or sell high-value, pre-built skills.

AI Training

I Built 4 AI eCommerce Agents Making $10K+/Month (n8n Templates)

The Recap: In this video, I'm gonna break down four different ecommerce AI agents and AI automations you can build, adapt, and sell to DTC brands and ecommerce store owners. We're going to cover systems that generate ad creative, handle social media monitoring + automatic replies, and creating influencer assets that feature your product.

P.S. We also launched a free AI Automation Community for those looking to build and sell AI Automations — Come join us!

The AI Gold Rush

The Recap: TSMC, the world's top chipmaker, just reported a record-breaking 39% profit surge, fueled by relentless demand for its advanced AI chips from tech giants like Nvidia and Apple.

Unpacked:

• The company's high-performance computing division, which builds chips for AI, generated a majority 57% of its third-quarter sales.

• Its most advanced chips, those smaller than 7-nanometers, now account for 74% of its revenue, showing the market's hunger for processing power.

• Demand is so strong that TSMC raised its 2025 revenue growth forecast to the mid-30% range, signaling confidence in the continued AI boom.

Bottom line: TSMC’s earnings are a powerful indicator of the massive hardware investment currently underpinning the entire AI industry. The unyielding demand for high-end chips confirms that the race to build more powerful AI tools is accelerating.

Where AI Experts Share Their Best Work

Join our Free AI Automation Community

Join our FREE community AI Automation Mastery — where entrepreneurs, AI builders, and AI agency owners share templates, solve problems together, and learn from each other's wins (and mistakes).

What makes our community different:

• Real peer support from people building actual AI businesses

• Complete access to download our automation library of battle-tested n8n templates

• Collaborate and problem-solve with AI experts when you get stuck

Dive into our course materials, collaborate with experienced builders, and turn automation challenges into shared wins. Join here (completely free).

The Robots.txt Rebellion

The Recap: Cloudflare is spearheading a publisher revolt against Google's AI Overviews, which are blamed for siphoning off web traffic. The company has rolled out a new Content Signals Policy to millions of websites, using the humble robots.txt file to legally challenge how Google uses content for AI summaries.

Unpacked:

• The move comes after a Pew Research study found AI Overviews nearly halved publisher referrals, dropping link clicks from 15% to just 8% on search pages with summaries.

• The new standard allows websites to separately permit crawling for traditional search while disallowing it for AI training and providing AI-generated answers.

• Google maintains that overall click volume remains stable and that AI features are driving higher quality clicks to websites.

Bottom line: This isn't just a technical update; it's a high-stakes attempt to redefine the web's unwritten rules for the AI era. The outcome could determine whether content creators are compensated or sidelined as AI becomes the new front door to the internet.

The Shortlist

GZDoom splintered after its lead maintainer inserted ChatGPT-generated code, prompting a group of developers to fork the popular open-source project into a new version called UZDoom in protest.

A former Biden energy envoy warned that the AI industry's energy demand is at risk of creating a "crisis," stating the current supply chain and power grid are unprepared for the scale of new data center clusters being planned.

Alibaba reported that its massive investments in e-commerce AI are already breaking even, with preliminary tests showing a 12% increase in return on advertising spend from its new tools.

OpenAI needs to spend over $1.3 trillion by 2030 to deliver on its promised 26 gigawatts of computing capacity, according to a Citi analysis, raising questions about its ability to fund the massive infrastructure buildout.

Signing off,

David, Lucas, Mitchell — The Recap editorial team