You're pledging to donate if the project hits its minimum goal and gets approved. If not, your funds will be returned.
A series of in-person hackathons on copilots for (conceptual alignment) research, with a specific flavor of design philosophy.
Here’s a blurb that hints at some of the specific flavor:
AI can work with unstructured data directly, instead of projecting them down to low-dimensional formal representations that machines have required pre-AI. What might the future of “postformal” interfaces look like, that takes this to a deep level? Where humans primarily supply the 0 to 1, the relevance and vision, and AI helps actuate them from 1 to 100? And for sensemaking the AI future specifically? To explore this, we’re going to do experiments in interface tech and culture, for collaborating with the AI that will be integrated in our infrastructure.
The paradigm shift invited by the near future, is moving away from scaling by replicating fixed structure (think MacDonald's & Fordism) and to scaling by attentive infrastructure that comes up with personalized structure in milliseconds. Even if AI is only mildly creative for a few years, its wide adoption creates ontological changes, not just quantitative changes.
If you want some more concrete examples of such interfaces (+ vague slogans for “postformal” tooling and how it relates to AI alignment), you can read the more participant-facing FAQ-like explainer here. Highly recommended if you like weird approaches to AI alignment.
Some goals:
Validate some of the hypotheses around live theory. In brief, “live theory” is about the new adaptive theories (powered by more short-term AI tech) that will be necessary to sensemake the crazy future (from extreme AI sophistication).
Experiment and build some sensemaking tools that will be actually useful already, and chart out some concreteness within this strange design philosophy.
Partner with established conceptual alignment researchers and foster collaboration with (research) engineers.
Explore opportunity modeling, to be properly responsible with dreaming a safe and beautiful future.
Help upcoming AI safety researchers meet the rigorous thinking demanded in AI risk and craft plans that don’t suck.
Experiments in wholesomeness/non-abusive tech in response to a particular threat model of tools that are not just frustrating but abusive.
Generate some very straightforward, real-world engineering+cultural output as stepping stones to a new organization based on the above philosophy.
We have already run the hack0 alpha version, we will be holding one more next weekend and then one more in CEEALAR, Blackpool, UK
https://docs.google.com/spreadsheets/d/1SD2SNZiYS6Z03-yn1zs1QaADRhwl5Sv6QqCd9q3SSXk/edit
The spreadsheet above details the breakdown of costs for renting the venue, buying snacks and food for participants, prizes for the winners, API credits for participants, infra costs for posters, marketers, whiteboards, flights cost to travel to UK for the 3rd iteration and salary for organisers.
We just ran the 0th version of this hackathon series! We had 27 sign ups, 5 submissions and 2 judges over 2.5 days. All this with barely any marketing.
We are compiling a report and we might be able to share the submissions once we figure out IP stuff.
The main team consists of Sahil (formerly SERI MATS, MIRI, currently founding an AI safety org) and Aditya (AI safety community builder in India, IISc PhD student).
Sahil has a strong conceptual research + engineering background with experience in AI risk research and skill and experience in facilitation and community building.
Aditya has been working with the Community Builders Grant team of CEA for more than a year now focusing on AI safety field building. He has technical expertise in machine learning studying at India’s top fundamental research university and has given talks at various IITs and IIITs on risks from general intelligence and at EAG London on community building. [previously manifunded]
There are many people contributing in some way or another to the project. We will specify names once we get permission to share.
Both the chances and downside of any serious “failure” is low here, since this is just a hackathon. However, for completeness’ sake:
Failure could look like:
It has to be canceled because of logistical issues
No one shows up, potentially because of time-crunch and marketing difficulties (already falsified given the amazing response we received)
People show up, and it is too chaotic to be a focused effort, or participants are misled or their time is wasted. (Also falsified for the weekend length)
Someone creates a raw capabilities start-up, or tooling that feeds into capabilities that competes with safety consideration
Causes could be:
Visa issues for traveling to venue
Poor organization or preparation
CEEALAR has offered to fund the venue and food for a week long workshop in November 2024.