Pro-social Agent Engineer
About the Foresight Institute:
Foresight Institute is a research organization and non-profit that supports the beneficial development of high-impact technologies. Since our founding in 1986 on a vision of guiding powerful technologies, we have continued to evolve into a many-armed organization that focuses on several fields of science and technology that are too ambitious for legacy institutions to support. From molecular nanotechnology, to brain-computer interfaces, space exploration, cryptocommerce, and AI, Foresight gathers leading minds to advance research and accelerate progress toward flourishing futures.
Create a forkable virtual world filled with pro-social AI agents that help humans reach their goals while respecting the basic rights of meat.
Although this sounds like a daunting task, it can be broken down into three steps:
The Learn phase is fairly straightforward and technically feasible today. We just need to create Digital Twins of real-world entities and populate them with data.
These entities primarily fall into two categories:
- organizations – nonprofits, governments, companies, etc.
- people – scientists, employees, doctors, patients, etc.
Each of these digital twins should have its own vector database of defining data. The most fundamental components of each entity would be its:
- wishes – goals, desires, objectives, etc.
- resources – data, humans, capital, etc. that can be allocated, pooled, or exchanged with other entities to fulfill wishes.
These digital twins serve primarily as an easy source of information for other agents (deciders and actors).
- Planning Agents that create hypotheses about the most efficient ways to fulfill various wishes while respecting universal human rights. These Decider Agents then use semi-quantitative cost benefit analysis to decide which strategies seem most promising.
- Humans can then review these strategies and decide which ones to test. Once a strategy is selected, the Planning Agents can then create Actor Agents to carry out the strategies.
Actor Agents then carry out plans and measure the results. Given current their limitations, Actor Agents should:
- automate what they can, using tools
- outsource what they can’t to humans by creating bounties
- Create a pro-social agent that improves the world using an agent framework and submit a pull request.
- Since reasoning capabilities are still limited, you might want to focus on the first “Learn” phase and create a Digital Twin of an organization or person.
- As a possible first step, we suggest training data ingestion and curation pipeline (Digital Twin Safe) described here.
- AI Agent Development: Create and refine proof-of-concept AI agents using existing autonomous agent frameworks (e.g., AutoGen, TaskWeaver, ChatDev, MetaGPT). Ensure these agents can simulate real-world scenarios and decisions for optimal societal outcomes.
- API & Framework Expertise: Develop robust APIs and work with OpenAPI specifications. Utilize frameworks such as Next.js and React for scalable, efficient applications.
- Documentation & Support: Produce comprehensive documentation to assist users and contributors in understanding and utilizing AI agents.
- Community Engagement: Organize and lead supportive events and meetings for contributors, fostering a collaborative environment.
- Project Management: Regularly update on project progress, strategize future developments, and navigate potential obstacles. Coordinate with various organizations and contributors for seamless integration and progress.
- Hackathon Coordination: Plan and execute hackathons focusing on specific autonomous agent frameworks, guiding participants in creating diverse AI agents for public good projects.
Skills and Experience
- Technical Expertise: Proven experience in creating APIs, OpenAPI specifications, and working with frameworks like Next.js and React.
- AI and Machine Learning: Experience in fine-tuning and deploying open-source AI models, including large language models. Familiarity with autonomous agent frameworks such as AutoGen, TaskWeaver, ChatDev, and MetaGPT.
- Open Source Contributions: A history of significant open-source contributions, showcasing your skills and collaborative nature.
- Communication and Leadership: Strong communication skills for effective documentation, meetings, and collaborative projects. Leadership experience in guiding teams and managing complex projects.
- Ethical AI and Data Privacy: Knowledge of ethical AI practices and data privacy regulations to ensure responsible development and deployment.
- Problem-Solving and Adaptability: Ability to navigate and solve complex technical challenges in a rapidly evolving field.
- Passion for Social Impact: A deep interest in using technology for societal betterment, aligning with the vision outlined in Gaming the Future.
About the role
This position is a project-based contractor role and is fully remote.
The compensation is $90,000 for the completion of the full project, paid according to milestones delivered.
Please apply by filling out this form.