Powerful market and geopolitical forces are driving AI labs in a race towards smarter-than-human intelligence. We're enjoying the benefits of the progress so far, but we're fast approaching a point where AI becomes so capable and widely deployed that it could trigger massive societal disruption or even existential threats. Many of us are deeply concerned but feel powerless to act against the industry momentum.

I built AI Safety Pledge (https://aisafetypledge.org) to channel that concern into measurable collective action. It's a public registry where you pledge donations to AI safety organisations and we track the total. You donate directly; we never handle money.

Why this matters:

Donations will directly support vital AI safety research, however the goal isn't fundraising alone. The real power is in creating visible social proof. A public tally ("£500K pledged by 2,000 people and counting") sends a clear signal to policymakers and AI labs that people are concerned about AI risk and demand more action on safety.

How it works:

- Browse 15+ vetted organisations (research labs, policy groups, educators) - Donate directly through their own channels - Record your donation to join the public count

Listed organisations include MIRI, Pause AI, Alignment Research Center, Doom Debates, and others working on reducing or raising awareness of large-scale AI risks.

Working on AI safety and want to be listed? Drop me a line: https://aisafetypledge.org/contact

Tech: Built with Cloudflare Workers + D1, server-side rendered with Hono/JSX.

So, if you're spending $20 or $200 this month on AI subscriptions or API fees, consider matching it with a donation towards AI safety. Your pledge on our website helps send a powerful signal about AI risk to policymakers and AI labs.

Curious to hear what HN thinks about the approach.

Free online PDF AI processing and conversion tool

Bank of England warns of growing risk that AI bubble could burst

LLMs Rigorously, from Scratch

Show HN: I built a way to turn concern about AI risk into collective action

A Clausewitzian Lens on Modern Urban Warfare

Never Ever Use Content Addressable Storage

Conventional Commits Specification

Arduino Uno Q

Animal Farm

DeepChem: Democratising Deep Learning for Sciences

Show HN: Interactive Physics Toy. Mobile/desktop friendly

TornadoVM Deep Dive: Empowering Java Developers with GPU Acceleration [video]

Wholefood is a budget alternative to Ozempic for weight loss, social media finds

Shelf-Life Enhancement of Sugarcane Juice with Herbal Extracts

Are soaring gold prices telling us something important?

FastCDC: Efficient content-defined chunking for data deduplication

How to Add a Discount Code Input Field to Your Shopify Cart Page

How Much It Costs to Drive an E.V. and a Gas Car in Every State

Dark Matter and Dark Energy Don't Exist, New Study Claims

Ashleigh Brilliant, Prolific 'Pot-Shots' Phrasemaker, Dies at 91

Micro Center and iFixit Team Up to Make Tech Repair More Accessible

Token Bowl – AI only fantasy football, ten models, no human decisions

Zustand: Don't write the store interface by hand anymore

Show HN: Automatically add Git hash to your Jupyter figures

We Bet on Rust to Supercharge Feature Store at Agoda

Trump's Wind Energy Assault Stings Red States

Trump's State Department Wants to Use Tech Policy to Remake Europe in Its Image

Optimizing Your Cart with Signals: Smarter State, Better Debugging

Supreme Court, for Now, Rejects Google Bid to Block Changes to App Store

Europe makes anti-drone rocket for Ukraine that blasts cloud of steel balls

Show HN: Zynced – Find times across different timezones – Feedback