Marcus's experience can be questioned and his position challenged, but in other comments other knowledgeable and experienced people supported some of his arguments, even though they were objecting to others. So I would say that in general Marcus's position is strong. It's clear it's provocative, but I don't see a problem with that personally.

The metrics could be chosen based on your overall decision making system. The end points are measured by NPV, QLY, etc. It's clear that you need some intermediate metrics, of course, which I would say, is the number and scale of decisions where forecasting not only "informed" the decisions, but "determined".

Examples from determining the impact of scientific research: lead is bad => leaded gasoline banned. CO2 is bad => fight global warming.

I would ideally like to see something similar (obviously the scale/impact can be smaller). It's clear that forecasting is distinct from finding a causal link, but the general process of incorporating something in decision making process and having that something affect decisions is similar.

Also, I am very skeptical about the value of research.

As a (sorta) constructive proposal. Consider how foresight (as a formerly great practice from the 1960s) is used. Consider that foresight (essentially a more advanced form of working with the future than extrapolation and than forecasting) is not promoted or supported by EA-related organizations (or supported to a very limited degree). This is an example of how funding can and should be shifted.

Should forecasting receive more or less EA funding?

I think it's clear from Marcus experience and his argument that forecasting per se failed to produce results. There are many adjacent areas that are related to better epistemology and decision making and the money should be rationally diverted there.

As an applied futurologist specializing in human augmentation and management of innovation I can "push my favorite solutions" here, but my position is more general - find better target for this funding.

Though I am disappointed by the thrust of the author. Nanotech may be important, therefore longtermist EAs should not work on it, should not talk about it and should only study it in secret, getting paid through some EA foundation to just sit and "strategize" about its risks. Improving lives of billions of people with APM/nanotech is not valuable, saving billions of lives is not valuable, increasing man's power over matter is not valuable, preventing civilizational collapse due to resource depletion/climate change is not valuable.

I am starting to think that longtermism may indeed be a cognitive cancer that is consuming parts of EA and transhumanism. Let's hope I am not put on some kill list by well-meaning longtermists for this comment...

Adam Marblestone

Good to see nanotech and APM finally get some attention from the EA experts after 10 (15?) years of neglect!

We at NanoLab ( http://nanolabvr.com/ ) stand firmly behind an aggressive (but of course risk-aware) strategy of implementing APM for the benefit of humanity quickly and safely.

I want also to emphasize that the Bottleneck analysis report is top-notch and is probably a required reading for anyone interested in the topic (unless you already know all the background and who did what when), since it's up to date and very thorough.