I attended this talk not knowing much about mechanistic interpretability at all and came away quite excited about the idea of working on it. Particularly, I found that there were concepts and intuitions around MI that overlap or have similarities with fluid mechanics and turbulence, which were the focus of my PhD. This surprised me and I've since been looking into MI further as something I could work on in the future. 

I also think there could be similar transferable intuitions from other fields of physical engineering which I'm interested in exploring further to help other engineers transition into the field (as part of my work at High Impact Engineers).

Thanks for giving this talk and sharing such a comprehensive write-up, Neel!

This is a fantastic overview, thanks for sharing it! 
There are a few more 'Groups by profession' orgs in this post. (Although, the specifics of some of the groups may be out of date now). 
 

Fantastic, I'll check those out, thanks!

Have you looked into integrating any of these chatbots with Emacs through APIs?

Thanks for a great post. 

+1 to the following:

Often by the time projects receive funding they don’t know “what to do with the money” and start looking fast into fiscal sponsors or other ways to receive the funds

Specifically, this is a concern for projects that have more than one paid member as one person receiving the money for many people and then distributing it can have adverse tax implications for the person receiving the large lump sum. We were in this position last year, along with quite a few (5+) other orgs we were communicating with. Happy to discuss this more if it's helpful.
 

Physical engineering (i.e. not software engineering) is another option if you're interested in a STEM field. For some examples of projects you could work on with an engineering skillset you can look at https://forum.effectivealtruism.org/posts/Bd7K4XCg4BGEaSetp/biosecurity-needs-engineers-and-materials-scientists . I'm not sure if it's the best undergraduate degree for research in biorisk but I think it's a valid option for biorisk more generally.

(Disclaimer: I help run High Impact Engineers)

Thanks for sharing this, I found it really interesting!

Are there any resources in particular (e.g. 1-3) that you would recommend for learning more about this topic?

Civilisational resilience is a cause area that we're definitely looking to add! If you've got any recommended starting points for engineers to learn more about the area, I'd love to know.

That's a fantastic outcome, congratulations!

Feedback on the data analysis: I find the 90% confidence interval to not be very informative in some situations. For example in the case of "Computing the ratio of money raised to time spent, we arrive at an average of 786 USD per hour (CI 7 USD to 3,100 USD)", the CI range is very broad and I only really take away 'there was high variance in the data' or 'there weren't many data points' or both. In this situation, it's both and it's relatively clear to see from the figures. 

I'm unsure what would be more informative. Maybe a box plot of the direct work hourly rate and your hourly rate calculations could be a good way to visualise the uncertainty. A statistical test to compare the means of the two distributions (direct work and fundraising) could also be interesting but maybe there's too much variance and too few data points in the fundraising data for now. 

It would also be interesting to see the maximum and minimum money raised per hour of time values.

Regardless, it's a great result and I look forward to hearing how the 2022 season goes!

Thanks for the elaboration, Michael. Pleased to hear you like the new name - your feedback was significant in the decision  to debate a name change, so thank you for providing it.  Your phrasing generally describes the broader discussion we had on this argument for the name change during our deliberations - it's a good reference for others who are considering this argument.

We had an interesting point come out of our deliberation on this topic of "apparent monopolisation": we want to address a lack of community for physical engineers in EA so, in a way, we want to monopolise the space so as to collect as many engineers as possible in one place to maximise network effects. This initially updated us relatively strongly towards keeping "EA" in the name.  However,  a counterpoint is that having a name that is sufficiently recognisable to EAs - "High Impact [x]", for example - hopefully derives a decent portion of the monopolisation benefit, while reducing the risk of "apparent monopolisation". Therefore, we concluded that the "maximise network effects by monopolisation through using EA in the title" point only weakly updated us towards keeping "EA" in the name. 

We would be very excited to see more organisations doing things for physical engineers in the future. We've been bouncing around quite a few ideas as we've been developing a strategy for the org, most of which will be out of our scope. So we would be interested in discussing ideas with anyone who is interested in starting an org in this space!

(I'll note that we discussed this in some detail and that the use of 'we' and 'us' in this comment may not accurately reflect the views each team member but hopefully I've captured our average view relatively accurately)