The bots are coming 🤖

Plus, Google goes for vibes

Welcome back! An AI-generated Val Kilmer is starring in a new movie. The late actor will appear in As Deep as the Grave as a Catholic priest, created entirely through generative AI with the blessing of his estate. 

Kilmer had signed on years before his death, but his throat cancer made filming impossible. It's the first time a posthumous performance has been fully AI-generated rather than CGI face-swapping. As studios and actors battle over AI likeness rights, this may be a preview of what's to come.

Google Launches Stitch, an AI-Native Design Canvas

Via YouTube

Google is betting big on … vibes. The company announced that Stitch, its design tool from Google Labs, is evolving into a full AI-native design canvas—one where you can turn natural language into high-fidelity UI without touching a wireframe.

What’s new: Instead of starting with mockups, you describe what you're building—your business objective, what you want users to feel, or examples that inspire you. Stitch generates designs from there. 

  • Add context: You can bring images, text, or code directly to the canvas as examples. 

  • Iterate: A new agent manager lets you work in multiple directions simultaneously without losing track.

  • Switch between apps: Stitch can now extract design systems from any URL or export them via a new markdown file called DESIGN.md, making it easier to move between Stitch and other tools. 

  • Interact instantly: Click "Play" to preview your app flow, and Stitch will auto-generate logical next screens based on user clicks. Google also introduced voice capabilities, so you can speak directly to the canvas and get real-time critiques or updates.

The bigger picture:"Vibe coding" has been the phrase of the moment—building software by describing what you want rather than writing every line. Now Google is pushing "vibe design" into the mix. The promise: close the gap and shorten the timeline from idea to reality.

Will this democratize design or flood the world with more AI slop? 

Cloudflare CEO: Bot Traffic Will Exceed Human Traffic By 2027

The bots are coming! In an interview at SXSW this week, Cloudflare CEO Matthew Prince predicted that AI bot traffic will exceed human traffic online by next year.

Prince's prediction: Before generative AI, the internet was only about 20% bot traffic—mostly Google's web crawler and a handful of bad actors. But AI has changed the math. When a human shops for a camera, they might visit five websites. An AI agent doing the same task might visit 5,000. Multiply that across millions of queries, and you get an explosion of traffic growing faster than anything since the early days of Covid streaming—except this time, there's no plateau in sight.

What’s next: The scale of what's coming is hard to fully grasp. Prince said the shift will require new infrastructure, like sandboxes for AI agents that spin up on the fly and tear down when tasks finish. It will also demand enhanced monitoring to keep things safe as bots become the dominant presence online. My two cents? We're heading toward a web built for machines first, humans second.

Pentagon Says Anthropic Poses 'Unacceptable Risk To National Security'

The Department of Defense fired back at Anthropic this week, calling the AI lab an "unacceptable risk to national security" in its first formal rebuttal to Anthropic's lawsuits.

What DOD is arguing: In a 40-page court filing, the Pentagon said it's concerned Anthropic might "attempt to disable its technology or preemptively alter the behavior of its model" during warfighting operations if the company feels its "red lines" are being crossed. The dispute stems from Anthropic's refusal to let AI be used for mass surveillance of Americans or in targeting decisions for lethal weapons.

  • Legal experts say the Pentagon's argument relies on speculation rather than evidence. "The government is relying completely on conjectural, speculative imaginings to justify a very, very serious legal step," one former Justice Department attorney told TechCrunch.

What happens next: Anthropic has requested the court temporarily block the DOD from enforcing its supply-chain risk designation. A hearing on that preliminary injunction is set for next Tuesday.

The bigger picture: The two sides remain at an impasse. Tech companies including OpenAI, Google, and Microsoft employees have filed briefs in support of Anthropic, arguing the Pentagon could have simply ended the contract rather than labeling the company a national security threat. How this gets resolved—and what it means for AI companies negotiating with the government—is anyone's guess.

AI search is exploding, and your customers are already asking ChatGPT, Perplexity, Claude, and Gemini who to trust. How do you make sure your brand is the answer? Enter: Peec AI.

Peec AI leverages AI search as a measurable growth channel and lets you:

  • Track your brand’s visibility and sentiment across major LLMs

  • Benchmark your share of voice against competitors

  • Follow step-by-step guidance to improve AI visibility

Ready to turn AI search into real growth?

Automated client reporting without the busywork

Via ViewMetrics

ViewMetrics consolidates data from Google Analytics, social ads, and email platforms into reusable Google Slides and Sheets templates. 

How you can use it

  • Pull data from multiple marketing platforms into one dashboard

  • Create reusable templates that auto-update on your schedule

  • Embed charts and visuals directly into Notion or client decks

  • Skip the manual export-and-paste routine every reporting cycle

Pricing: Free and paid plans

A voice-activated AI agent for your Mac

Lemon turns natural spoken instructions into finished tasks across your apps—drafting emails, creating documents, researching, and taking actions without switching tabs.

How you can use it

  • Dictate emails and have Lemon draft, reply, or send on your behalf

  • Create documents and notes without touching your keyboard

  • Research topics and get summaries while you keep working

  • Delegate repetitive tasks across apps with a single voice command

Pricing: Free (early-stage, expect evolving features)

Jobs, announcements, and big ideas

  • DoorDash tests a new “Tasks” platform that pays drivers to gather real-world data for training AI and robotics systems.

  • OpenAI deploys GPT-5.4 Thinking to oversee internal coding agents and catch alignment issues in real time.

  • Jeff Bezos is looking to raise $100 billion for an AI manufacturing fund.

  • Uber commits $1.25 billion to Rivian in a fresh partnership aimed at scaling robotaxi development.

  • CodeRabbit introduces Plan, a new layer designed to align coding agents on intent before execution begins.

  • Lovable broadens its platform beyond app creation with new capabilities for file analysis and multimodal content generation.

Is Meta replacing creators with AI? Here’s what stood out from this week in AI.

That’s a wrap! See you next week for more. Enjoy your weekend.

—Matt (FutureTools.io)