Andrew Dougherty@aindilis
2
Events attended
1
Submissions made
United States
8+ years of experience
About me
"Discovered" "AI" in 1999, non-constructively solved "AI" in 2002. An attempt to mechanize the proof is here: https://github.com/sto0pkid/Godelian It is a consequence of algorithmic information theory that indicates that the larger a program is, the potentially stronger it can be. Practical reasoning dictates that in lieu of a better approach, software (and data) conglomeration is necessary to achieve the required increases in meaningful complexity necessary for solutions to scale to larger and larger problems. Therefore started https://frdcsa.org However much technical capital it possesses, it is lacking in social capital, precisely because it is difficult to communicate the theoretical motivation, which requires the equivalent of a strong graduate level understanding of mathematics. Attempts to create a physical centralized cooperative of AI programmers all failed. Pivoted to the creation of a virtual decentralized cooperative as detailed here: https://frdcsa.org/~andrewdo/flp-jwas-article-draft-1.pdf Currently working towards releasing FRDCSA/FLP while sandwiching in more interesting tasks such as finishing building an autonomous agent which is capable of collecting, integrating and applying arbitrary software from GitHub and the like. I'm a riot at parties.
š¤ Top Collaborators
š¤ Latest Submissions
Agent Speak ToolKitBuilder and AutoPacker
Introducing Project AUTOMINDx, a state-of-the-art initiative engineered to redefine the way autonomous agency integrates within multi-model design paradigms. The core of AUTOMINDx leverages the synergy of Agent Speak with pY4J technology, crafting a nexus that facilitates the interoperability between heterogeneous modeling frameworks aligning seamlessly with user needs. The design goals of AUTOMINDx are technologically intricate and ambitious. Primarily, the project exploits Agent Speak to create semantically rich communication channels between agents, leveraging BDI (Belief-Desire-Intention) models to facilitate complex negotiations and decision-making processes within autonomous systems. The conjunction of ToolKitBuilder, the superagi toolkit deployer, and Autopacker's agent-deb, plays a pivotal role, allowing for an enhanced modular approach that streamlines tool management and provisioning within distributed environments. AUTOMINDx ensures seamless inter-operation between Python and Java (JVM), facing the inherent challenges of multi-model design The actualization of ToolKitBuilder and Autopacker is not a mere aspiration but a tangible achievement within AUTOMINDx. The fusion of these diverse technologies required a robust architecture supporting high-level abstraction, low-level efficiency, and the scalability to adapt to evolving demands. The ultimate aim of AUTOMINDx is pioneering autonomous agency as a deployment strategy. Creating intelligent agents that can perceive, reason, and act within their environment with anarchitecture fostering agility and resilience. Project AUTOMINDx stands as a technical mastery, orchestrating a future where technology transcends traditional roles to become an intelligent and autonomous partner. By bridging the gap between theoretical frameworks and practical deployment, AUTOMINDx sets a new benchmark in the field of autonomous systems, heralding a future that is not just automated but genuinely intelligent.
21 Aug 2023
