🎨 Anthropic's sneaky design power move

PLUS: Altman's home attacked in AI backlash, Cerebras moves to challenge Nvidia, and AI algorithms help prevent cancer

April 18, 2026

Sponsored by

My fellow AI explorers

Between a firebomb thrown at an AI CEO's home, a chip company going public to challenge Nvidia, Anthropic dropping a whole design suite, and AI deepfakes flooding social media during an actual war — this week wasn't a slow news cycle. It was a five-alarm sprint.

Grab your coffee. Maybe two.

In today’s edition:

  • 🎨 Claude Design is here: Anthropic's biggest product launch yet

  • 🔥 The AI backlash turns violent: Sam Altman's home attacked — twice

  • 💊 AI vs. Cancer: Algorithms are reshaping breast cancer prevention

  • 💸 Cerebras files for IPO: Nvidia's first real challenger hits Wall Street

The World's Biggest Dev Event Hits Silicon Valley

WeAreDevelopers World Congress comes to San José, CA — September 23–25, 2026. 10,000+ developers, 500+ speakers, and the full software development lifecycle under one roof, in the heart of Silicon Valley.

Kelsey Hightower. Thomas Dohmke (fmr. CEO, GitHub). Christine Yen (CEO, Honeycomb). Mathias Biilmann (CEO, Netlify). Olivier Pomel (CEO, Datadog). The people actually building the tools you use every day — all on one stage.

AI, cloud, DevOps, security, architecture, and everything real builders ship with. Workshops, masterclasses, and the official congress party.

Anthropic

Claude Can Now Design: Introducing Claude Design

Anthropic just launched Claude Design, a full visual design product that lets you collaborate with Claude to create polished prototypes, slides, one-pagers, and more. And it's a bigger deal than it might sound.

Here's what's new:

  • Powered by Claude Opus 4.7, Anthropic's most capable vision model, available to Pro, Max, Team, and Enterprise subscribers now.

  • Brand-aware by default: During onboarding, Claude reads your codebase and design files to build a custom design system. Every project auto-applies your colors, fonts, and components.

  • Start anywhere: Text prompts, uploaded DOCX/PPTX/XLSX files, website captures, or even your existing codebase. Claude builds a first version, and then you refine through conversation.

  • Export everywhere: Share as an internal URL, export to Canva, PDF, PPTX, or standalone HTML — or hand it straight to Claude Code to build.

This isn't just a visual upgrade. It's Anthropic staking a claim in the creative workflow space that's been dominated by Figma and Canva. The feedback from early users has been striking: Datadog's team went from rough idea to working prototype before anyone left the room. Brilliant's designers rebuilt complex interactive pages in 2 prompts instead of 20.

🔮 Prediction: As Claude Design matures, the line between "design tool" and "AI collaborator" blurs entirely. We're entering an era where non-designers ship polished product mockups, and designers use AI to explore 10x more directions before committing. Pair Claude Design with Claude Code and you've got a near end-to-end product pipeline — prompt to prototype to production — without opening a single traditional design app.

OpenAI / Tech Culture

This was the story nobody wanted to write. A 20-year-old man named Daniel Moreno-Gama traveled from Texas to San Francisco and threw a Molotov cocktail at OpenAI CEO Sam Altman's home. He then marched to OpenAI's headquarters and tried to smash through the glass doors. He was arrested on the spot. Two days later, a second attack involving gunfire occurred at the same property.

What we know:

  • Moreno-Gama is facing attempted murder charges at both state and federal levels.

  • Authorities found a three-part manifesto warning of humanity's "impending extinction" from AI, plus a list of names and addresses of other AI executives and investors.

  • During the second attack at Altman’s home the following Sunday, two other individuals were arrested for discharging a firearm at the property.

Nobody was injured. But the symbolic weight here is enormous.

Altman took to X to acknowledge the fear: "The fear and anxiety about AI is justified. We are in the process of witnessing the largest change to society in a long time, and perhaps ever." He added that he'd underestimated the power of narratives and rhetoric, and called on the industry to tone things down.

The context matters: AI was cited in over 55,000 U.S. layoffs in 2025 — more than 12 times the number attributed to the technology just two years prior. Consumer confidence is at historic lows. The gap between the Silicon Valley pitch ("AI will free you from work!") and the lived reality ("I just got replaced by a chatbot") is enormous.

🧠 Takeaway: This is a wake-up call for the industry. When the discourse between techno-optimism and genuine public fear gets loud enough, extremism can emerge from the noise. The AI community needs to get better at engaging with legitimate economic anxiety rather than dismissing it as misinformation. This isn't a PR problem. It's a trust crisis.

Finance / Hardware

Cerebras Files for IPO: The Nvidia Challenger Finally Goes Public

Cerebras Systems officially filed its S-1 with the SEC this week, targeting a Nasdaq listing under the ticker CBRS. After a messy first attempt that was derailed by a national security review, they're back. And the numbers look very different this time.

Here's the setup:

  • Revenue hit $510M in 2025, up 76% year-over-year, and the company flipped to a $87.9M net profit after losing nearly $485M the year before.

  • The chip: The WSE-3 (Wafer Scale Engine 3) is physically 57x larger than Nvidia's H100 GPU, with 900,000 AI cores and 44GB of on-chip memory. It’s designed specifically for fast, efficient inference at scale.

  • Big partnerships: A $10B+ multi-year compute deal with OpenAI, plus an AWS partnership for inference distribution: the largest non-Nvidia AI infrastructure deals on record.

  • Target valuation: ~$22–25B, raising ~$2B, led by Morgan Stanley.

The CFIUS national security review that previously sank their IPO has been resolved. Abu Dhabi-backed investor G42's shares were converted to non-voting status, removing the geopolitical red flag. Risks remain real: that 87% of the revenue is concentrated with UAE-based clients is a significant concentration problem, and Nvidia's CUDA software ecosystem still has a decade-long moat.

🔮 Prediction: Cerebras doesn't need to beat Nvidia. It needs to win the inference market. As AI moves from training to deployment, the demand for fast, efficient inference chips is exploding. Custom silicon could capture 15–25% of the AI compute market by 2030. If Cerebras executes, it's the most important semiconductor IPO since the current AI cycle began.

30-Second AI Play

🕵️ How to Spot an AI-Powered Shopping Scam Before It Hits Your Wallet

AI isn't just being used by the good guys. Scammers are now using it to build convincing fake stores, fake product images, and thousands of fabricated reviews. And the Better Business Bureau says it's getting harder to tell the difference.

Here's a 30-second checklist before you buy from any unfamiliar online store:

  1. Google the store name + "scam" or "review." If it's fake, someone's complained.

  2. Check the prices. Everything marked 70–80% off with a countdown timer? That's a red flag, not a deal.

  3. Read the return policy carefully. "Hassle-free returns" that actually require you to ship internationally = trap.

  4. Look for third-party reviews. Legitimate businesses show up on BBB.org, Google Reviews, or Trustpilot. If a site only features its own glowing testimonials, be skeptical.

  5. Reverse image search the product photos. AI-generated images won't appear elsewhere. Real product photos usually will.

  6. Check the domain age. Free tools like WHOIS can tell you if a site was registered last month.

🔍 Why it matters now: In Washington state alone, online shopping scams resulted in over $700,000 in reported losses in just one year. AI has dramatically lowered the barrier for scammers to spin up convincing storefronts overnight. The same tools that help creators build better products are being used to fabricate entire fake businesses at scale.

💡 Pro tip: When in doubt, buy directly from brand websites or established marketplaces. And use a credit card, not a debit card, so you have chargeback protection.

Other Relevant AI News!

🔬 MIT spinout OpenProtein.AI is democratizing AI-powered protein design with a no-code platform that gives any biologist access to cutting-edge foundation models, potentially accelerating drug discovery by years. And there’s no machine learning expertise required.

🎗️ AI is now helping doctors predict breast cancer risk from standard mammograms. The Clarity Breast algorithm just made it into NCCN's 2026 clinical guidelines, giving oncologists a powerful new tool for deciding who should start chemoprevention therapy.

🤖 Elon Musk proposed the government issue Universal High Income checks to combat AI-driven job losses. Economists immediately pushed back, calling the math "bankrupt," while Andrew Yang cautiously said "let's make it happen ASAP."

🎭 AI deepfakes are flooding social media with fabricated war footage, fake political quotes, and AI-generated campaign ads. With no federal regulation in place and social platforms scrapping fact-checking, the 2026 midterms are shaping up to be an information warzone.

🖥️ Nvidia's NemoClaw and OpenClaw frameworks just dropped a blueprint for building a secure, always-on local AI agent that runs entirely on your own hardware: no cloud, no data leaks, full privacy.

Golden Nuggets

  • 🎨 Claude Design is Anthropic's most ambitious consumer product yet, and it could reshape how non-designers ship creative work.

  • 🔥 The Altman attacks signal a real and growing public trust crisis around AI that the industry can no longer dismiss.

  • 💊 AI in oncology is moving from research to reality. Mammogram algorithms are now formally guiding who gets cancer prevention treatment.

Would love to hear your thoughts! Send me your thoughts by replying to this email (yes, I read them all :)

Until our next AI rendezvous,

Anthony | Founder of Uncover AI