Forethought[1] is a new AI macrostrategy research group cofounded by Max Dalton, Will MacAskill, Tom Davidson, and Amrit Sidhu-Brar.
We are trying to figure out how to navigate the (potentially rapid) transition to a world with superintelligent AI systems. We aim to tackle the most important questions we can find, unrestricted by the current Overton window.
More details on our website.
Why we exist
We think that AGI might come soon (say, modal timelines to mostly-automated AI R&D in the next 2-8 years), and might significantly accelerate technological progress, leading to many different challenges. We don’t yet have a good understanding of what this change might look like or how to navigate it. Society is not prepared.
Moreover, we want the world to not just avoid catastrophe: we want to reach a really great future. We think about what this might be like (incorporating moral uncertainty), and what we can do, now, to build towards a good future.
Like all projects, this started out with a plethora of Google docs. We ran a series of seminars to explore the ideas further, and that cascaded into an organization.
This area of work feels to us like the early days of EA: we’re exploring unusual, neglected ideas, and finding research progress surprisingly tractable. And while we start out with (literally) galaxy-brained schemes, they often ground out into fairly specific and concrete ideas about what should happen next. Of course, we’re bringing principles like scope sensitivity, impartiality, etc to our thinking, and we think that these issues urgently need more morally dedicated and thoughtful people working on them.
Research
Research agendas
We are currently pursuing the following perspectives:
* Preparing for the intelligence explosion: If AI drives explosive growth there will be an enormous number of challenges we have to face. In addition to misalignment risk and biorisk, this potentially includes: how to govern the development of new weapons of mass destr
1. Get a pilot up and running NOW, even if it's extremely small.
You will cringe at this suggestion, and think that it's impossible to test your vision without a budget. Everyone does this at first, before realizing that it's extremely difficult to stand out from the crowd without one. For you, maybe this is a single class delivered in a communal area. 30 students attending regularly, demonstrating a good rate of progress, is a really compelling piece of evidence that you can run a school.
- Do you have the resilience and organisation skills it takes to independently run a project?
- Will people actually use it?
- Can you keep your staff?
- Can you cost-effectively produce results?
It can compelling prove the above, whilst having a ton of other benefits.
2. YOU need to be talking to funders NOW
Don't fall into the trap of trying to read their minds. Get conversations with them. Get their take on your idea. Ask what their biggest concerns would be. Go address them. Repeat. Build relationships with them and get feedback on your grant proposals before submitting them.
As the founder, its YOUR job to raise money. Don't delegate it. It'll take forever to get them to understand your organisation well enough, they won't be as sufficiently motivated to perform, and you won't learn. This is going to be a long-term battle that you face every year. You need to build the network, skills & knowledge to do it well.
3. Be lean AF
The best way to have money is not to spend it. Both you and your charity may go without funding for months or years. Spend what little money you have, as a person and as a charity, very slowly. The longer you've been actively serving users, the easier fundraising gets. It's about surviving until that point.
4. Funders will stalk your website, LinkedIn, and social media if they can
As much as possible, make sure that they all tell the same story as your grant application - especially the facts and figures.
5. When writing your proposals, focus on clarity and concreteness above all else
Bear the curse-of-knowledge in mind when writing. Never submit anything without first verifying other people can understand it clearly. Write as though you're trying to inform, not persuade.
- Avoid abstractions
- State exact values ("few" -> "four", "lots" -> "nine", "soon" -> "by the 15th March 2024")
- Avoid adjectives and qualifiers. Nobody cares about your opinions.
- Use language that paints a clear, unambiguous image to the readers mind
OLD: mean student satisfaction ratings have increased greatly increased since programs began and we believe it's quite reasonable to extrapolate due to our other student-engagement enhancements underway and thus forecast an even greater increase by the end of the year"
NEW: When students were asked to rate their lessons out of 10, the average response was 5. Now, just three months later, the average is 7/10. Our goal is to hit 9/10 by 2025 by [X,Y,Z].
Good luck!