New Technology / Ai Agents
Internet of Cognition
Track AI agents, autonomous workflows, agentic software tools and real-world adoption signals across the next wave of AI products.
Source material: Scaling Intelligence Out: Cisco's Vision for the Internet of Cognition, with Vijoy Pandey
Key insights
- Vijoy Pandey advocates for an Internet of Cognition where AI agents collaborate and share context, potentially transforming problem-solving in shared environments
- The current focus on scaling individual AI models risks centralizing power, highlighting the need for distributed systems that promote wider participation and control
- Ciscos Community AI Platform Engineer (CAPE) exemplifies distributed AI by employing 20 agents to automate complex cloud tasks, enhancing efficiency and user response times
- The AGNTCY project seeks to create an open-source framework for AI agents to communicate and collaborate, essential for fostering diverse AI interactions
- A healthcare demo showcased how multiple agents can collaborate, demonstrating the practical advantages of AI in improving service delivery
- Pandeys vision promotes a stable AI architecture that emphasizes permissioned participation and transparency, aiming to create safer and more accountable systems
Perspectives
Analysis of the Internet of Cognition and its implications for AI collaboration.
Vijoy Pandey's Vision
- Proposes the Internet of Cognition for AI agents to collaborate effectively
- Highlights the need for distributed systems to enhance AI capabilities
- Emphasizes the importance of secure and observable connectivity
- Advocates for task-based access control to improve security
- Describes the Community AI Platform Engineer (CAPE) as a multi-agent system improving operational efficiency
- Introduces cognition engines as accelerators or guardrails for AI interactions
Concerns and Challenges
- Questions the effectiveness of decentralized identity systems for accountability
- Raises concerns about the unpredictability of agent behavior
- Highlights the potential for miscommunication among agents
- Notes the challenges of ensuring agents revert to base-level access after tasks
- Warns about the risks of specialization leading to siloed knowledge
- Expresses skepticism about agents ability to maintain shared understanding
Neutral / Shared
- Acknowledges the need for robust frameworks for agent discovery and communication
- Recognizes the importance of observability in multi-agent systems
- Notes the potential for AI to enhance productivity in various sectors
- Mentions the role of open-source projects in advancing AI collaboration
- Discusses the evolving nature of AI technologies and their implications
Metrics
automation
40%
percentage of tasks automated by the Community AI Platform Engineer
This level of automation significantly reduces the workload on site reliability engineers.
this system has reduced load on site reliability engineers and improved response times for end users by fully automating some 40% of tasks.
other
four core pillars in its businesses
Cisco's business focus areas
Understanding these pillars helps in grasping Cisco's strategic direction.
it's networking, it's security, it's observability, and it's collaboration.
other
the OSI seven-layer model
Framework for networking protocols
This model is essential for understanding how different technologies interact.
there is this OSI seven layer model for networking.
other
Ethernet is probably something that most people are familiar with
Common networking technology
Ethernet's role is crucial in local area networking.
Ethernet is probably something that most people are familiar with.
other
traffic predictions
application of machine learning in network systems
It allows for foresight in managing network loads during significant events.
traffic predictions because you can foresee events
other
noise reduction
application of machine learning in WebEx
It improves user experience during virtual meetings.
WebEx uses ML pipelines to do things like noise reduction
workload_reduction
30%
reduction in team workload due to automation
This reduction indicates significant efficiency gains for the team.
we reduced the load on the team by 30%
task_automation
40%
percentage of tasks automated by the Cape platform
Automating tasks allows teams to focus on more complex issues.
40% of the tasks that the team handles have actually been agentified
Key entities
Timeline highlights
00:00–05:00
Vijoy Pandey emphasizes the importance of an 'Internet of Cognition' for AI agents to collaborate effectively, which could revolutionize problem-solving. Cisco's Community AI Platform Engineer demonstrates the potential of distributed AI systems to enhance efficiency in complex tasks.
- Vijoy Pandey advocates for an Internet of Cognition where AI agents collaborate and share context, potentially transforming problem-solving in shared environments
- The current focus on scaling individual AI models risks centralizing power, highlighting the need for distributed systems that promote wider participation and control
- Ciscos Community AI Platform Engineer (CAPE) exemplifies distributed AI by employing 20 agents to automate complex cloud tasks, enhancing efficiency and user response times
- The AGNTCY project seeks to create an open-source framework for AI agents to communicate and collaborate, essential for fostering diverse AI interactions
- A healthcare demo showcased how multiple agents can collaborate, demonstrating the practical advantages of AI in improving service delivery
- Pandeys vision promotes a stable AI architecture that emphasizes permissioned participation and transparency, aiming to create safer and more accountable systems
05:00–10:00
Cisco emphasizes the importance of secure and observable connectivity for collaboration among digital entities. The company advocates for distributed systems to enhance flexibility and decentralization in AI and networking.
- Cisco aims to provide secure and observable connectivity across platforms, which is essential for collaboration among people, machines, and objects in the digital age
- The company promotes distributed systems for horizontal scaling, enhancing flexibility and robustness in computing to manage complex workloads effectively
- Vijoy Pandey stresses the need for decentralization in AI to counter the concentration of power among a few leading companies, advocating for a participatory internet
- The OSI seven-layer model serves as a framework for understanding networking protocols, clarifying the functionality of technologies like Ethernet and TCP/IP
- Pandey notes the increasing role of machine learning in network management, raising concerns about the balance between human-designed protocols and automated systems
- The conversation highlights the necessity of a foundational layer that supports decentralization, ensuring the internet remains open for participation and innovation
10:00–15:00
The integration of machine learning in network management is enhancing data flow and issue resolution. Cisco's applications of machine learning, including anomaly detection and traffic prediction, are improving operational efficiency.
- The integration of machine learning in network management is advancing, emphasizing the importance of data flow and issue resolution for future operations
- Current networking relies on deterministic processes and established protocols, which are crucial for ensuring reliable connections and data integrity
- Control plane software is essential for managing network policies and routing, particularly during outages, as it enables real-time traffic adaptation
- Machine learning applications in network systems, such as anomaly detection and traffic prediction, improve operational efficiency and help prevent failures
- Ciscos use of machine learning spans various applications, including customer sentiment analysis and tools like WebEx, highlighting its role in enhancing user experience
- Projects like Jarvis represent a move towards more sophisticated automated solutions in network management, addressing the need for intelligent systems to optimize operations
15:00–20:00
The Jarvis project utilizes coding agents to enhance site reliability engineering, significantly improving operational efficiency. The Cape platform automates 40% of tasks and reduces team workload by 30%, showcasing the benefits of AI integration in operational frameworks.
- The Jarvis project applies coding agents to site reliability engineering, improving operational efficiency and reflecting collaborative problem-solving in software development
- Cape, a community AI platform engineer, automates the SRE pipeline, reducing team workload by 30% and automating 40% of tasks, which enhances response times
- Major companies like Adobe, AWS, and Nike collaborate on the Cape project, promoting cloud-native operational excellence and driving innovation in multi-agent systems
- Efficiency improvements from Jarvis benefit both the SRE team and the developer community, highlighting the value of AI integration in operational frameworks
- The open-source nature of the Cape project fosters wider participation and innovation in AI-driven operations, enabling continuous adaptation to technological challenges
- Advancements in agentification and automation at Cisco indicate a shift towards more intelligent infrastructure management, essential for addressing modern software demands
20:00–25:00
American companies are increasingly remaining private, limiting investment opportunities for everyday citizens. VCX by Fundrise aims to democratize access to private tech investments, including AI and space sectors.
- American companies have historically led global innovation, but the trend of remaining private has limited investment opportunities for everyday citizens. VCX by Fundrise aims to democratize access to private tech investments, including AI and space sectors, for those previously excluded
- Current multi-agent systems struggle with complex tasks autonomously, highlighting the need for human expertise in technology-driven environments. Understanding these limitations is essential for shaping future technological roles
- The pursuit of Artificial Superintelligence (ASI) focuses on both vertical and horizontal scaling, with the latter offering untapped potential for collective intelligence. This shift could redefine machine intelligence and enhance technological capabilities
- The vision for ASI includes a team of agents that can innovate independently of human input, representing a significant advancement in technology. Achieving this would transform the landscape of machine intelligence and its applications
- The industry needs to prioritize collaboration among AI models rather than just improving individual ones. This collective approach could lead to significant breakthroughs in problem-solving and innovation
25:00–30:00
Achieving advanced artificial superintelligence necessitates a focus on horizontal scaling of intelligence, emphasizing the potential for collective intelligence among agents. The evolution of AI development mirrors human intelligence, suggesting that enhanced collaboration among AI systems is crucial for addressing complex challenges.
- Achieving advanced artificial superintelligence requires a focus on horizontal scaling of intelligence, as current methods overlook the potential for collective intelligence among agents
- AI performance should be evaluated based on how effectively a well-organized team of humans can complete tasks, highlighting the importance of collaboration among AI systems
- The development of language significantly accelerated human intelligence, suggesting that a similar breakthrough in communication among AI agents could enhance their collaboration and task delegation
- As AI agents advance, they must align their intents and collaborate autonomously on complex tasks, which is crucial for addressing unprecedented challenges
- The evolution of AI development parallels human intelligence, indicating that smarter agents will need to learn collaborative skills to unlock new problem-solving capabilities
- The future of AI collaboration envisions a richer interaction among agents, moving beyond simple task delegation to foster a diverse AI culture across various contexts