Grants for creators who want to explore the risks of AI
AI is moving faster than humanity’s ability to control it. We support YouTube creators to make long-form videos raising awareness of catastrophic AI risks, in their own style.
.jpg)
I also run Species, a YouTube channel about AI risks that averages around 1.5 million views a video. I’m extremely worried about where AI is headed. That’s why I’m working with the Center for AI Safety, a nonprofit research organization, to get more creators making videos about AI dangers.
Why YouTubers? Most people have no idea what the latest AI can do. Voters and politicians can’t react to a danger they don’t know exists. Long-form YouTube is where a lot of people now go to understand big topics. I want more people outside the tech industry to learn what’s at stake and have a say, because otherwise a few tech companies will decide for us.
An AI experiments channel walks through real research on AI agents that escape their sandboxes and fight to survive.
A professor’s game theory channel analyzes why wars start, and what that means for a possible conflict between humanity and AI.
A film analysis channel revisits classic sci-fi to see how its AI warnings compare with the risks we face today.
A philosophy channel looks at a real experiment where 1,000 AI agents formed their own society in Minecraft.
We agree on the topic with you up front, then read the script and the finished video once to check facts and confirm it’s still on topic.
Each video includes a spoken credit of about 15 seconds saying CAIS supported it and pointing viewers to slowdown.ai, with our logo on screen. We also require a credit and link in the description.
We’d like to support several videos with each creator, as many as suit your channel. A single video is welcome too.
Our YouTube strategists can give notes on titles, thumbnails and scripts, and we can connect you with AI researchers to check facts.
This program is for creators who take these risks seriously and want to help more people understand them. They need to be a substantial part of the video, but the tone and format are up to you. We welcome creators from all backgrounds and niches.
People could use AI to engineer new pandemics, build chemical weapons, or launch cyberattacks on critical infrastructure. Autonomous AI weapons could let conflicts spiral out of control.
Countries and companies are racing to build the most powerful AI. Each is afraid of falling behind, so they keep pushing ahead and cutting corners on safety, raising the risk of catastrophic outcomes.
Governments could exploit AI to spread propaganda, quell dissent, and infringe civil liberties, enabling an oppressive concentration of power.
Future advanced AIs might “go rogue” and pursue goals counter to our interests, evading our attempts to redirect or deactivate them. They might become power-seeking and engage in deception.
We review every application. If your channel is a good fit, we’ll be in touch with next steps.
The Center for AI Safety (CAIS) is a nonprofit whose mission is to reduce societal-scale risks from artificial intelligence. Alongside our research, we help the public understand those risks, and these grants are part of that effort. You can find out more about our work on the About page.
The Statement on AI Risk, published by CAIS in 2023 and signed by hundreds of AI scientists and public figures, including Geoffrey Hinton and Yoshua Bengio.
