What is this blog about?

This blog, Forecasting AI Futures, is as it sounds about forecasting AI-related things. I am focusing on things that I perceive as having high relevance to AI safety, regarding strategy and preparation for highly advanced intelligence.

My hope and goal for this blog is to fill an important gap in current forecasting efforts by leveraging prediction markets and making probabilistic forecasts about a wide range of strategically relevant questions.

Prediction markets, as well as Metaculus, are powerful forecasting tools, achieving accurate predictions by aggregating the beliefs of many forecasters. I will support my analyses using predictions from such platforms, with the additional benefit that it is easy to check how the predictions change over time.

Here are some things I plan to write about, to give you a taste:

  • AI-Related Disasters: When might they occur, and how severe could they be?

  • International Coordination: How likely is it that an AI-induced disaster could lead to the formation of treaties or international regulatory bodies?

  • Development Conditions for Artificial General Intelligence (AGI): Under what conditions will AGI emerge? Will it be developed in a lightly regulated environment, under strict governmental control, or in adherence to international agreements?

  • Societal Reactions: How swiftly and intensely will people respond to advances in AI? Will there be significant anti-AI protests or acts of terrorism?

If you find probabilistic predictions for questions like these useful, this blog is for you.

Why subscribe?

To be notified when I publish new posts of course!

Also, for a spam-free, ad-free reading experience, plus audio and community features, get the Substack app.

Why Get a Paid Subscription?

To support this blog.

Since I am writing things that I hope will improve people’s ability to prepare for and improve the future, I don’t want to paywall the posts. But I still need some income to be able to focus on this. I am hoping that some kind souls will take it upon themselves to make it possible for me to continue writing about the future.

I might make a few posts only for paid subscribers, particularly regarding specialized topics that not everyone might find useful. I will also prioritize answering comments and emails from paid subscribers.

A founding member will get the opportunity to write guest posts, if they want to. I will of course review any guest post before publishing, mainly to check that it is relevant enough to the theme of this blog.

User's avatar

Subscribe to Forecasting AI Futures

A forecasting blog for AI safety relevant predictions. Exploring likely futures to support strategic decision making and preparation for highly advanced AI.

People

Author of the Forecasting AI Futures blog. Constantly thinking about the potentials and risks of future AI systems.