Emergence

Welcome to Emergence, where the future of enterprise workflow automation begins. Listen to exciting words from our co-founders Satya Nitta, Sharad Sundararajan, and Ravi Kokku, Learn Capital's founder and investor Rob Hutter, our research scientist Ashish Jagmohan, and our Chief Design Officer Hélène Alonso as they share how we’re advancing the science and development of #AIagents. Follow us to discover how intelligent agents will unlock the full potential of #AI in enterprise systems.

Three papers to read on enhancing multimodal interactions, written by Emergence’s engineers: 1. The Picture Says It All! Multimodal Interactions and Interaction Metadata ➡ https://lnkd.in/e6w3SBCU 2. Robust User Context Analysis for Multimodal Interfaces ➡ https://lnkd.in/evnUVq2F 3. Designing Multiuser Multimodal Gestural Interactions for the Living Room ➡ https://lnkd.in/ejTjje-R #AI #AIagents #MultimodalInteractions

For truly efficient programming systems, having some of the base-level attributes of agents as superclass templates is crucial. Not only does this enable developers to extract from these and repeatedly reuse concepts rapidly, but it also maintains consistency in the definition and programming of agents. There is one such example of an agent communication and collaboration paradigm that has greatly eased the building of multi-agent collaboration systems (i.e., AutoGen). However, many more base characteristics of agents need to be standardized and frameworks built to enable large-scale agent-oriented programming paradigms. Some of these are illustrated below. The inclusion of security templates is a base-level requirement, ensuring adherence to constitution. Similarly, self-improvement is a pivotal component of any agent system. However, it is important to ensure that self-improvement stays true to some properties of goodness and the agent’s constitution. Read more about the Agent-Oriented Programming Paradigm here 👉 https://lnkd.in/e4ctp8qC #AI #AIagents #AOP #AgentOrientedProgramming

The present Agentic era is marked by the explosion and proliferation of agents, and it's important to look back at the recurring historical patterns to effectively manage these systems in the near future. Evolution in the past advanced from monolithic to complex distributed architectures, necessitating advanced orchestration and routing. This aided scalability, which in turn advanced interoperability, composability, reusability and, most importantly, discoverability. The figure below provides a clear breakdown of each era. The personal computing era witnessed a shift from centralized mainframes to distributed computing, which was managed through innovations like TCP/IP routing. This acceleration advanced in the Internet era as static websites and monolithic servers transitioned to dynamic server clusters and microservices, requiring sophisticated service routing techniques. The cloud computing era brought further decentralization with distributed databases, enhancing data management through advanced slicing and transaction routing. In the current #Agentic era, there is a need to orchestrate LLM-based distributed agents. This significantly allows for improving task specialization, optimizing resources, continuous learning and collaborative problem-solving. By directing tasks to specialized agents best suited to solve these complex issues, orchestration improves both efficiency and output quality, thereby harnessing each agent's strengths and compensating for weaknesses. Agents, thereby, work in synergy and adaptively learn from interactions, fine-tuning strategies over time. Orchestration enables dynamic adjustment of task routing and resource allocation based on real-time feedback and maintains high performance amid changing conditions. It also effectively breaks down complex problems into manageable components, utilising collaborative efforts to provide innovative solutions. The inherent feedback loops in orchestration increase the system's robustness and resilience by refining system behaviors and fostering the development of new operational dynamics. Read more about the evolution of systems here: https://lnkd.in/eq-WYxxy #AIagents #AI

Our Senior Machine Learning Engineer, Aakash Nain, released a new paper on the flow of information in a pre-trained transformer, in collaboration with Sakana AI. Read it now 👇

With increasing interest in #AIagents and systems, a myriad of definitions have emerged over time, especially after the introduction of large language models (#LLMs) and vision language models (#VLMs). As we advance this science, it is useful to have one encompassing definition that is extensible as well as expressive, building on commonly understood concepts from the past. Emergence's definition of an agent refers back to a few basic principles that embody an agent, using them to develop the notion of an “agent object” in the world of LLMs and LVMs. For an agent object to be programmed scalably to build robust systems, we believe it must embody the following primary concepts: ✔️ Autonomy: An agent is an autonomous, interactive, goal-driven entity with its own state, behavior, and decision-making capabilities. It has the capability of self-improvement when it sees that it is unable to meet the performance parameters for reaching its goal. ✔️ Reactivity and Proactivity: Agents can be reactive (understand their environment and respond to changes using actions) and be proactive (take initiative based on their goals). Agents can take actions that change the state of their environment, which can ensure their progress towards their goal. ✔️ Beliefs, Desires, and Intentions (BDI): A common model used in agent-oriented programming (AOP) is the BDI model, where agents are characterized by their beliefs (information about the world), desires (goals or objectives), and intentions (plans of action). ✔️ Social Ability & Communication: Agents have a communication mechanism and can interact with other agents or entities in their environment. This interaction can be highly complex, as it may involve negotiation, coordination, and cooperation. We believe that with the advent of LLMs, this communication can be completely based on natural language, which is both human- and machine-understandable. ✔️ Constitution: An agent needs to adhere to some regulations and policies depending on the imperatives of its task and goals. It needs to protect itself from being compromised or destroyed as well as be trusted not to harm other agents sharing the environment in which it is operating. ✔️ Memory: An agent must retain Long Term Memory (LTM) to ensure the successful completion of a task. When successfully employed by an agent, it can significantly reduce repetitive computing and reproduce human demonstrations. It also has short-term memory, which relies on any information based on the available context length. Learn more about the anatomy of #AI agents here 👉 https://lnkd.in/e4ctp8qC

“Emergence” is a compelling phenomenon observable both in natural systems and in engineered designs, where complex behaviors and patterns arise from simple interactions. As James McClelland and John Holland have illustrated, systems composed of simple agents can evolve to exhibit intricate patterns and capabilities that transcend those of any individual component. For example, imagine millions of birds following just three simple rules: avoid crowding your neighbors (separation), fly in the same direction as those around you (alignment), and stay close to the flock (cohesion). Nothing in any individual bird’s “programming” calls for the complex murmuration for which starlings are known, pictured here. Yet, when many starlings form a system, their repeated basic interactions give rise to this breathtaking behavior. This is emergence in action – the birth of complex phenomena from surprisingly simple, non-linear rules. Read more here: https://lnkd.in/eq-WYxxy

In his book The Society of Mind, Marvin Minsky conceptualizes the mind as a society of small components with specialized goals and functions, otherwise known as agents. Organized hierarchically, each agent is responsible for different mental functions and has its own rules for interacting with other agents. These agents possess the ability to work both individually and collectively as a group, advancing cooperation and leading to intelligent behaviour and cognitive processes through their interaction. As such, no individual agent is built to form consciousness or intelligence—these phenomena emerge naturally from the system of agents. This is what, today, is a cornerstone of our ideology. Learn more in our recent blog post, ‘The Emergence of Emergence’ 👉 https://lnkd.in/eq-WYxxy #AI #AIagents

Our parent company doing great things with really smart people! #generativeAI #AIagent

"Some of the world's greatest AI systems have been touched by people who work here," said Satya Nitta, our CEO and Co-founder. At Emergence, we are pushing the boundaries of AI with a team that embodies diversity, talent, and expertise. Our greatest strength lies in the exceptional individuals who drive our mission forward. Our team is composed of experts from some of the world’s leading #AILabs, including IBM Research, Google Brain, Alexa, the Allen Institute of AI, Meta, and Microsoft. This wealth of experience equips us to build scalable and enduring systems. We collectively have 25+ PhDs and over 1,000 patents. But we are not just engineers or developers; we are architects of cognitive ecosystems, crafting tools that think, learn, and ultimately, understand. At the core of our development is a commitment to social intelligence — the idea that our technologies should enhance the capabilities of human teams and networks. We are dedicated to shaping the future of #GenAI, enhancing workflows, and creating an innovative agentic future. Discover more about our team and projects here 👉 https://www.emergence.ai/ #AI #AIagents