MIT analyst details transition from generative to action-oriented AI agents
MIT associate professor Phillip Isola discusses the shift from generative AI to agentic systems that use software wrappers and tools to perform tasks. The interview highlights potential applications for streaming infrastructure such as automated coding and digital customer service, while noting the risks of human de-skilling and data limitations.
Key Takeaways
- 35% of surveyed businesses have already deployed AI agents as of November 2025
- Agentic systems differentiate from generative AI by taking autonomous actions like booking flights or managing files
- Missing training data for button-clicking and UI navigation remains the primary hurdle for digital agent development
- Coding agents are the most successful current application due to their ability to verify outputs via feedback loops
- Risks include 'vibe coding' where developers fail to verify agent-generated code, leading to bugs or data leaks
Why It Matters
The shift toward agentic AI marks a transition for streaming platforms from using AI as a creative assistant to an operational workforce. For B2B infrastructure providers, this means automating the most labor-intensive parts of the stack, including live code deployment and multi-layered customer support. As these agents evolve to process multimodal data beyond text, they will likely manage complex streaming workflows such as real-time ad inventory packaging and delivery optimization. Watch for the emergence of 'agentic ecosystems' where various specialized AI entities coordinate to maintain high-availability streaming pipelines without human intervention.
Additional Context
The move toward agentic systems is accelerating across the enterprise landscape. Per Gartner, August 2025, 40% of enterprise applications are expected to feature task-specific AI agents by 2026, a significant increase from less than 5% in 2025. This rapid adoption is driven by the evolution of foundation models like Gemini 3.1 Pro and GPT-5, which now feature advanced multi-step planning and autonomous decision-making capabilities. Gartner also forecasts that by 2028, a typical Fortune 500 company could manage more than 150,000 AI agents, highlighting a looming need for centralized governance to prevent 'agent sprawl.' In the media sector, major cloud and hardware providers are already rolling out infrastructure for these systems. Per NVIDIA, March 2025, the company launched AI Blueprints for video search and summarization, enabling agents to reason through live camera streams and provide real-time operational insights for large-scale facilities. Similarly, AWS demonstrated agentic media operations at its 2025 re:Invent conference, focusing on transforming manual orchestration steps into intelligent, agent-driven processes for live sports and multi-channel advertising. Streaming leaders are also integrating agentic logic into consumer-facing and back-end tools. Per TechRadar, July 2025 edition, Netflix use of AI for complex visual effects sequences, such as a collapsing building in 'The Eternaut,' demonstrates a shift toward more autonomous production tools. Meanwhile, Disney showcased its 'Select AI Engine' at CES 2025, which uses machine learning to autonomously expand audience reaches and automate ad campaign activations via its Disney Compass platform. These developments suggest that the near-term future of streaming will be defined by 'systems of action' that prioritize real-time intelligence and operational scalability.
Read full article at goodmenproject.com
Get this in your inbox → Subscribe
Enjoy our coverage?
Add StreamingMeme as a preferred source on Google to see more of our streaming news at the top of your Search results.
Add as preferred source