- Future Tools
- Posts
- The "Intelligence Age" is coming
The "Intelligence Age" is coming
Vibe coding frenzy

Welcome back! I'm at HumanX this week in San Francisco. There are literally hundreds of companies here, all focused on enterprise adoption and practical uses of AI. On Monday, it was great to be in the room for Fei-Fei Li's opening keynote on world models and the future of embodied AI.
If you're at the conference, come say hi! 👋

Hello from HumanX! Ft. Catherine McMillan from my friends at The AI Collective


OpenAI Releases Policy Recommendations for the ‘Intelligence Age’

Via WIRED
OpenAI is going beyond building AI. This week, the company released a policy document titled "Industrial Policy for the Intelligence Age," proposing a range of recommendations tied to AI and its potential to upend work and the economy.
What it's advocating for:
A public wealth fund that distributes cash to citizens, giving them "a stake in AI-driven economic growth"
Incentives for employers to experiment with four-day work weeks, as long as output doesn't fall
Active measurement of how AI affects wages and unemployment—with automatic triggers for increased unemployment benefits or job training once metrics cross thresholds
Faster electrical grid development to support AI infrastructure
Why now: OpenAI says the proposals are a "starting point" for discussion. Chief global affairs officer Chris Lehane told Bloomberg that policy conversations around AI need to be as transformative as the technology itself. "It is simply not good enough to wave your hands and say, 'Here's all the things that are going to happen' and then not actually come up with solutions," he said.
Why it matters: Being a frontier lab isn't just about building models anymore. It's about creating the broader societal response to what those models make possible. OpenAI is betting that getting ahead of the policy conversation is as important as winning the technical race.
Vibe Coding Is Flooding the App Store
A wave of new apps are launching in Apple’s App Store. The likely reason? Vibe coding. The Information reported that in the first quarter of this year, the number of new apps published globally was up 84% compared to the same period a year earlier. It’s a big change from previous years when new apps were down 48% between 2016 and 2024.
Who’s building what:
Vibe coding tools like Claude Code and Codex have made it possible for non-programmers to build working apps using written prompts—and for experienced developers to ship far more code than before.
The majority of new apps are in the productivity category, with photo, video, and weather apps also surging.
Replit users alone have published nearly 5,000 apps to the App Store in the last few months, despite Apple’s crackdown on the tool.
The bigger picture: Vibe coding means democratization—but more apps aren’t necessarily better. Developers and consumers are already complaining about low-quality apps flooding the store, making discovery harder. As one consultant told the Information: "There's many more apps but not necessarily more time to add them to your day."
Google Quietly Launches Offline AI Dictation App
Google on Monday quietly released "Google AI Edge Eloquent," a free offline-first dictation app for iOS that takes on tools like Wispr Flow, SuperWhisper, and Willow.
What it does: The app uses Gemma-based speech recognition models that run locally on your device. You dictate, see live transcription, and it automatically polishes the text, filtering out filler words like "um" and "ah." Options like "Key points," "Formal," "Short," and "Long" let you transform the output. You can also turn off cloud mode entirely for local-only processing, import keywords from Gmail, and search through your transcription history.
Why the quiet launch: Google released this as an experimental app without a big announcement—likely testing the waters before rolling out improved transcription features more broadly. The App Store description hints at an Android version with system-wide keyboard integration and a floating button for easy access.
The bigger picture: AI-powered transcription apps are gaining serious traction as speech-to-text models improve. Google is clearly watching—and getting in on—the trend.


AI-powered ad inspiration

Via Hooksy
Hooksy indexes over 10 million social ads and uses AI to extract video transcripts, detect top-performing hooks, and generate scroll-stopping headlines on demand.
How you can use it
Build organized swipe files from winning ads across platforms
Spot creative trends without hours of manual research
Generate ad hooks and headlines on demand
Track competitor campaigns and repurpose content faster
Pricing: Paid

Secure handoff layer for AI agents

Via Kubbi
Kubbi creates temporary, encrypted claim links to pass sensitive payloads between agents or humans.
How you can use it
Pass one-time tokens, configs, or secrets between AI agents securely
Integrate into workflows via MCP/A2A tools, Python/TypeScript SDKs, or REST API
Enforce AES-256 encryption with automatic key rotation
Control exactly when and how payloads are retrieved
Pricing: Free and paid plans


Jobs, announcements, and big ideas
Anthropic introduces its Claude Mythos system card and restricts access to select defensive cybersecurity partners.
Spotify upgrades its AI playlist feature to generate mixes that now include podcasts.
OpenAI launches Prism, an AI-powered workflow designed to improve scientific paper review and reproducibility.
CapCut launches Dreamina Seedance 2.0 in the US with a free trial and steeply discounted Pro access.
The Central Intelligence Agency deploys a quantum sensing tool to locate a downed airman in Iran.
A new model called HappyHorse-1.0 unexpectedly takes the lead in text-to-video and image-to-video benchmarks.
World Labs releases upgraded Marble 1.1 and Marble 1.1-Plus models.


Anthropic reveals its secrets. Follow along as I walk through the Claude leak and everything else you need to know in AI right now.

That’s a wrap! See you Friday for more.
—Matt (FutureTools.io)