Long description

Context and background

SingularityNET Foundation, in collaboration with other partners such as the OpenCog Foundation and TrueAGI, is working toward a scalable implementation of the Hyperon AGI framework running on decentralized infrastructure, and toward implementation of the PRIMUS cognitive architecture within this framework.

Hyperon and PRIMUS are complex systems involving multiple components, which need to demonstrate appropriate functionalities both individually and in combination.  

This RFP aims to address a portion of this overall need, via funding the initial iteration of one significant component of PRIMUS within Hyperon.

The goal of this project is to develop a modular and extensible framework that supports a variety of motivational systems, including but not limited to Psi and OpenPsi. This framework should enable the seamless integration of these systems within AGI architectures like PRIMUS (formerly OpenCog) Hyperon, supporting not only human-like motivations but also the potential for “alien digital” intelligences with non-human motivational structures.

The framework must dynamically adjust motivational priorities in response to internal states and external stimuli, facilitating adaptive behavior in complex and unpredictable environments. Additionally, it should allow for alignment with overarching human ethical guidelines, embedding mechanisms to guide AGI behavior in accordance with human values and socially beneficial outcomes. Applicability within PRIMUS Hyperon infrastructure is the intent, ensuring the framework can be used to guide the development of and/or be implemented within core components such as ECAN (Economic Attention Allocation Network), DAS (Distributed Atomspace), and the MeTTa language.

As this is exploratory work, we want to be open to different ways of approaching it. For instance, we are open to either of:

1. a conceptual and/or mathematical specification of a detailed motivational framework.  In this case it should be precisely enough specified to guide implementation by an AI developer without expertise in motivational systems; or,
2. a simple, limited-scope code prototype of a motivational system, functioning in some limited domains and contexts, along with a clear plan regarding how to generalize and extend it to deal with broader situations.

To be clear, a proposal submitted for this RFP does not need to propose a full, scalable implementation of a motivational system for a PRIMUS system or other Hyperon-based HLAI. Rather, it should propose to do some preliminary work that would form the foundation for this sort of full scalable implementation (which may potentially then be funded by a future DF RFP).

Collaboration

This RFP will be followed by subsequent RFPs for applications that leverage Hyperon/PRIMUS to carry out various applications, and that aim to guide Hyperon/PRIMUS systems in cognitive development toward beneficial AGI

RFP expected outcomes

• Flexible motivational framework

• • Creation of a versatile framework that can integrate various motivational systems and adapt to different types of AGI, including human-like and alien digital intelligences.

• Detailed use cases

• • Provide a detailed explanation of how the given motivation framework would display itself in some concrete use-cases including: chatbot systems, and one or both of:
    • humanoid robots
    • or virtual agents / neoterics, such as in a metaverse environment like Sophiaverse

• Scalability and adaptability

• • A framework that seems likely to, in its complete and future implementation, be able to deal with large Atomspaces interacting with highly dynamically environments, dynamically adjusting motivational priorities based on internal and external factors.

• Ethical alignment

• • Consider exploring how AGI systems operate with motivational drives aligned with human values, fostering positive and socially beneficial behavior.

• Impact on Hyperon

• • The foundational work done here would, when more fully developed, lead to an integral framework to be used within Hyperon.  Such a framework would be expected to interact with several key elements of Hyperon such as:
    • ECAN (Economic Attention Allocation Network) Helping to allocate motivational priorities dynamically, influencing how ECAN manages resources and attention within the AGI.
    • DAS (Distributed Atomspace): Storing and managing the motivational states and decisions within the shared knowledge base.
    • MeTTa Language: The framework's logic and rules could be expressed and managed through MeTTa, allowing for flexible and complex motivation-driven behaviors within the AGI system.

• Foundation for future research

• • Establishment of a base for continued research and innovation in AGI motivation systems, paving the way for future advancements.