Topic Contributions


Why should we care about existential risk?

Here is my take on the value of extinction risk reduction, from some years ago: https://forum.effectivealtruism.org/posts/NfkEqssr7qDazTquW/the-expected-value-of-extinction-risk-reduction-is-positive

This posts also contains links to many other posts related to the topic.

Some other posts, that come to different conclusions:




One final thing: Generally, I think people make a distinction between existential risk (roughly: permanent, irreversible, and drastic loss of value of the future) and extinction risk (extinction of humans), where extinction risk is just one type of existential risk.

Samotsvety Nuclear Risk Forecasts — March 2022

Thanks so much for writing this!  I expect this will be quite useful for many people.

I actually spent some time this week worrying a bit about a nuclear attack on the UK, bought some preparation stuff, figured out where I would seek shelter or when I’d move to the countryside, and so on. One key thing is that it’s just so hard to know which probability to assign. Is it 1%? Then I should GTFO! Is it 0.001% Then I shouldn’t worry at all.

The Future Fund’s Project Ideas Competition

Enlightenment at scale (provocative title :-) )

Values and Reflective Processes (?), X-risk (?)

A strong meditation practice promises enticing benefits to the meditator---less suffering, more control over ones attention and awareness, more insight, more equanimity. Brahmavihara practice promises the cultivation of loving-kindness, compassion, and empathetic joy. The world would be a much better place if everybody suffered less, had more equanimity, and felt strong compassion and empathy with other beings. But meditation is hard! Becoming a skilled meditator, and reaping these benefits, requires probably thousands of hours of dedicated practice. Most people will just not put in this amount of effort. But maybe it doesn't need to be this way. The field of meditation teaching seems underdeveloped, and innovative methods that make use of technology (e.g. neurofeedback) seem largely unexplored. We are interested in supporting scalable solutions that bring the benefits of meditation to many people.


  • I don't actually know if meditation really has these benefits; this would needed to be established first (there should be quite some research on this by now). It seems plausible to me that meditation can be very beneficial. Several of my friends claim to have experienced significant benefits from meditation, and I think I can also point to tangible benefits in my own life.
  • These innovation need not be directly related to meditation; for example, one could imagine development of an extremely safe and non-addictive pharmaceutical substance that would let people experience, say, strong compassion, and thus increase compassion in everyday life (see e.g. the use of MDMA in therapy).
The Future Fund’s Project Ideas Competition

AI alignment prize suggestion: Improve our ability to evaluate (and provide training signal for) fuzzy tasks

Artificial Intelligence

There are many tasks that we want AI systems to do, for which performance cannot be evaluated automatically (and thus training signal provision is hard). If we don't make progress on our ability to train systems for such tasks, we might end up in a world full of systems that optimise for that which is easy to measure, rather than what we actually want. One example of such a task is the evaluation of free-form text; there is currently no automated method to evaluate free-form text (with respect to criteria such as usefulness or correctness) that matches human evaluation. The Future Fund could offer prizes for work that takes a task for which the gold-standard of evaluation is humans, and demonstrates an automated evaluation method that matches human evaluation very closely (or work that demonstrates an automated evaluation method to be superior to human evaluation).

Note: This is crucially not the same as "training models to perform well on the task in question". There are a number of technical reasons why what I suggest is easier. Intuitively, evaluating performance is often considerably easier than generating good performance. For example, I can watch a movie and say if it's good, but I can't make a good movie.

The Future Fund’s Project Ideas Competition

AI alignment prize suggestion: Demonstrate a true sandwiching project

Artificial Intelligence

Sandwiching projects are a concrete way for how to make progress on aligning narrowly superhuman models. They “sandwich” the model in between one set of humans which is less capable than it and another set of humans which is more capable than it at the fuzzy task in question, and b) figure out how to help the less-capable set of humans reproduce the judgments of the more-capable set of humans. For example, first fine-tune a coding model to write short functions solving simple puzzles using demonstrations and feedback collected from expert software engineers. Then try to match this performance using some process that can be implemented by people who don’t know how to code and/or couldn’t solve the puzzles themselves.

Importantly, there are many ways to attack a sandwiching project that are slightly cheating. The most challenging version of a sandwiching project would need to make sure that no information whatsoever from the more-capable set of humans is used in the training process. The Future Fund could offer prizes for demonstrations of sandwiching projects on various levels of impressiveness and generality of the employed method.

The Future Fund’s Project Ideas Competition

AI alignment prize suggestion: Introduce AI Safety concepts into the ML community

Artificial Intelligence

Recently, there have been several papers published at top ML conferences that introduced concepts from the AI safety community into the broader ML community. Such papers often define a problem, explain why it matters, sometimes formalise it, often include extensive experiments to showcase the problem, sometimes include some initial suggestions for remedies. Such papers are useful in several ways: they popularise AI alignment concepts, pave the way for further research,  and demonstrate that researchers can do alignment research while also publishing in top venues. A great example would be Optimal Policies Tend To Seek Power, published in NeurIPS. Future Fund could advertise prizes for any paper that gets published in a top ML/NLP/Computer Vision conference (from ML, that would be NeurIPS, ICML, and ICLR) and introduces a key concept of AI alignment.

The Future Fund’s Project Ideas Competition

Refinement of project idea #22, Prediction Markets


Add: "In particular, we'd like to see prediction platforms that do all of the following three: use real money, are very easy to use, allow very easy creation of markets.

The Future Fund’s Project Ideas Competition

Highly effective enhancement of productivity, health, and wellbeing for people in high-impact roles

Effective Altruism

When it comes to enhancement of productivity, health, and wellbeing, the EA community does not sufficiently utilise division of labour. Currently, community members need to obtain the relevant knowledge themselves and do related research, e.g. on health issues, themselves. We would like to see dedicated experts on these issues that offer optimal productivity, health, and wellbeing, as a service. As a vision, a person working in a high-impact role could book calls with highly trained nutrition specialists, exercise specialists, sleep specialists, personal coaches, mental trainers, GPs with sufficient time, and so on, increasing their work output by 50% while costing little time. This could involve innovative methods such as ML-enabled optimal experiment design to figure out which interventions work for each individual.

Note: Inspired by conversations with various people. I won't name them here because I don't want to ask for permission first, but will share the prize money with them if I win something.

Load More