The AI labs still have to earn our trust ... #AI #agents #OpenAI #productivity #futureofwork

YouTube · Watch/Read · (gathered 2026-09-23)

• AI labs must deliver reliably complex, “EA-sized” outcomes, such as fully planning a trip and managing conflicts, to earn user trust and justify a deeper, recurring relationship beyond simple assistive tasks.
• The content distinguishes between current basic AI assistance (e.g., surfacing a confirmation code) and future “EA-sized” work, which involves proactive planning, conflict resolution, and rule-checking.
• For AI agents to achieve this level of trust and integration, they must perform these complex tasks with “nearly 100% reliability,” as this is the critical bar for user adoption and willingness to pay.