Unforeseeable changes in my professional life have consumed the time and energy that I was previously dedicating to this blog and related efforts. While these changes are positive for me overall, I am struggling to even watch the spaces of forecasting and global catastrophic risk mitigation let alone contribute to them. More life changes on the horizon mean it’s unlikely I’m back in the next year or so, but I do hope to return to this and related projects eventually.

A brief retrospective…

I posted something to the blog every week for about 3 months. All of those posts represented meaningful work from my perspective, and I think each was either an effort to more clearly collect and communicate relevant information or an attempt to start building something new that I thought would be useful. I’m proud of this output for this timeframe, but obviously wish I could have sustained it for longer. I did not find the opportunities for collaboration on projects that I was initially hoping for, but I’m optimistic that these would have come with time.

My biggest cause of regret is not fulfilling the commitment I made in my proposal to continue this blogging effort until September. Two investors contributed a combined $1,000 based on that stated goal and the expected impact of my work. I’ll be reaching out to them individually to offer a refund of that investment.

While I’m away, if you find any of my work on the blog interesting and would like to discuss it, please still leave a comment on the post or reach out to me on Twitter. I’ll make time for a chat.

Thanks for reading.

Hi Austin. Thanks for the questions (and for building awesome tools in the forecasting world!). I just realized a few days ago, while listening to a podcast with Michael Aird, that I should probably be cross-posting at least most of the content to the EA forum. I plan on starting to work through the backlog very soon!

To answer the questions in order...

• Very little feedback so far, but this was about what I was expecting so early for such a niche topic. The EA forum might help me get feedback on my reasoning/writing/point me to new research sources, but I'd be pretty surprised if the posts got direct responses from people working on the problem of figuring out how to better forecast GCRs.

• The blog is actually focusing on the question of "how to improve the forecasting of GCRs" vs. me attempting to forecast GCRs. I have no reason to believe my current forecasts would be more accurate or useful than those that are already out there. My current mental model is that most of the impact will be from using the blog posts (specifically the "Application" ones that are sort of project outlines built off of the work in the rest) to do outreach to specific individuals and organizations in positions to either collaborate on implementing some version of them or to give feedback on why they don't think they're useful so I can create useful ones. I'm currently doing this for my piece on Open Philanthropy's internal forecasting, but I'm very excited about a piece I'm working on right now about how Metaculus could be connected to the EA forum to both reward posts that actually make people more accurate forecasters (tracked with citations in forecast rationales), and to incentivize the creation/concentration of deeper knowledge on forecasted topics that accumulates over time. I have a looser picture in my head of the next Application post, which will probably be how a Discord bot/servers could be used to create cultures of practice between forecasters and help them become more accurate by learning from each other.

P.S. - I'm currently pretty pessimistic about using prediction markets to forecast GCRs, mostly because of the incentives against information sharing and the long time horizons, but if you think differently, or think there's another promising direction that I'm missing, I'd love to chat sometime! If there is a possible solution in the prediction markets space, I'd expect it to be found in play money markets like Manifold.