AI-Powered Feeds and Networks Drive Unpredictable Traffic Patterns

YouTube has introduced a new feature that allows users to build custom feeds by describing the types of videos they want to see or recommend. This 'custom feed' feature uses Google's Gemini AI model to create a personalized feed based on the user's request.

The company also launched new capabilities for its 'Ask YouTube' feature, which enables users to ask multiple questions and receive organized video recommendations. For example, when searching for reviews of a specific product, users may receive a comparison table with product attributes and categories based on their preferences.

Cisco Systems released data today that shows how much capacity and bandwidth are required due to AI adoption rates. According to the report, 75% of companies have already deployed AI for network operations, while 51% run agentic AI systems. Cisco found that AI traffic is on pace to double every six months, with tasks performed by agents generating up to 450% more total network traffic when agentic AI is included.

Companies are increasingly turning to AI and agent-driven networks, with many organizations allowing AI agents to take on more responsibility in network operations. The report found that 82% of companies are comfortable granting AI a high or fully autonomous role, including making production changes without prior human approval.