AI & I Four Predictions for How AI Will Change Software in 2026
475 snips
Dec 31, 2025 
Brandon Gell, COO of Every, dives into fascinating predictions shaping software in 2026. He discusses the future of agent-native architectures, enabling AI to perform complex tasks seamlessly. Designers are poised to become powerful AI builders, leveraging creativity without technical barriers. The chat also highlights new engineering roles evolving from AI integration while exploring concerns like misinformation and the ethical implications of realistic AI-generated content. Gell emphasizes the need for tools that are accessible to all, paving the way for a more inclusive tech landscape.
AI Snips
Chapters
Transcript
Episode notes
Centralize Memory For Personalization
- Build a unified memory and context layer so every product can personalize itself to a user.
- Collect user data centrally so all apps and content can use the same context for smarter AI interactions.
Prioritize Agentable User Actions
- Make every app action executable by an internal agent so users can instruct the agent to act.
- Prioritize first-level agent capabilities (user actions) before deeper automation layers.
Agent-Native Architecture Emerges
- Agent-native architecture will make any action a user can perform delegable to an AI agent inside the app.
- This design level expands to app code and developer actions, enabling agents to modify and relaunch software.