newsfilter.io
Conference Presentation

Every CIO will have to answer for every token | Factory's Matan Grinberg

Sequoia CapitalMatan Grinberg
Watch on YouTube →
  • Distinct token limits for enterprise needs are expected to be addressed within approximately two years.
  • CIOs are projected to manage allocation strategies for every incremental token rather than relying on uniform limits assigned across roles.
  • Organizations may deploy fine-tuned models specifically for legacy code bases like COBOL instead of utilizing general-purpose models such as Opus.
  • Routing systems will accommodate varying organizational constraints by directing tasks like "vibe coding" to specialized models like Gemini Flash.
  • Reliability-focused workflows may integrate multi-model processes, such as generating code with OpenAI, testing with Anthropic, and reviewing with Gemini.