🤝In Partnership: Build the business you keep thinking about

🤿Deep Dive: Anthropic's Chris Olah called for outside voices to hold AI labs accountable

🤝Powered by: Join the biggest conference in the age of AI

🤿Deep Dive: Google Cloud's take on AI security

⚒Tool Snapshots: Tools for AI, no-code, and productivity

🖼AI Art: Examples of great and trending AI art

Olah was candid that even well-intentioned AI companies operate under commercial, geopolitical, and personal pressures that can quietly push them in the wrong direction, and that outside critics are essential precisely because of that.

He described AI models not as engineered systems but as something grown from human thought and language, more mysterious than their creators expected, with internal states that functionally resemble emotions like joy, fear, and grief.

He raised three urgent questions he believes the Church is uniquely positioned to push on: how AI's gains reach the global poor, what human flourishing actually looks like in an AI-saturated world, and what moral status these systems may eventually deserve.

On labor displacement, he warned that the harder challenge is not supporting those who lose jobs domestically but ensuring that wealth created by AI, concentrated in a handful of rich nations, finds its way to the rest of the world.

His closing ask was direct: more of the world needs to do what the Pope has done here, take AI seriously, look closely, and be willing to tell the labs when they are failing.

De Souza warned that security cannot be bolted on after the fact, and that companies need governance and auditability built into their AI platforms from the start, not left to individual employees to figure out.

He flagged "shadow AI" as a real risk, where staff reach for unsanctioned consumer tools, and stressed that any serious AI strategy has to be built alongside a data strategy and a security strategy.

His proposed answer is AI-native defense, where agents run security operations at machine speed under human oversight, and he called it a board-level issue, not just something for the security team.

While his advice is sound, Google itself is under scrutiny. Developers reported five-figure bills from unauthorized Gemini API calls after Google quietly expanded the scope of existing API keys without clear notice, and research found that even deleted keys can still be exploited for up to 23 minutes due to slow revocation across Google's infrastructure.

🧩 Unabyss - Keep your AI tools synced with your real context automatically.

🔗 own.page - Build a personal link page that actually feels like you.

⚙️ Yansu - Turn repetitive work into automations without starting from scratch.

📄 Tiny CV - Create polished resumes from markdown in minutes.

⌨️ Pi Coding Agent - Shape your coding agent around your workflow.