OpenAI cofounder Greg Brockman has acknowledged that the company's Plugins feature failed because AI models weren't sophisticated enough, despite being marketed confidently at launch. Looking forward, he envisions AI becoming an invisible layer that handles digital tasks autonomously, eliminating the need for users to learn software. However, current AI models lack the reliability needed for this vision, and businesses still require heavy custom integration work—a reality that has led OpenAI, Anthropic, and Microsoft to establish separate integration services.
Summaries like this, in your inbox every morning.
Sign up free →What happened
OpenAI cofounder Greg Brockman acknowledged that the company's Plugins product, launched in 2023 to add web search and third-party apps to ChatGPT, failed because AI models weren't ready. Despite this, OpenAI marketed the technology confidently at the time.
Why it matters
Brockman's vision is for AI to become an invisible interface—a persistent, context-aware agent that acts on its own rather than a feature-packed app. This represents a shift from current products like Codex, which still require users to learn software. However, AI models remain unreliable enough that implementation still requires heavy prompt engineering and custom integrations, which is why OpenAI, Anthropic, and Microsoft have all spun up separate companies to help businesses integrate AI.
What to watch
The gap between Brockman's stated vision and OpenAI's current products remains wide. AI models still aren't reliable enough to function as the autonomous, invisible layer he describes.
No comments yet. Be the first to share your thoughts!
Log in to join the discussion
Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.
Get Started FreeFree · takes 30 seconds · unsubscribe anytime
1 minute a day. The AI essentials.
200+ sources · Email / LINE / Slack
