Episode Summary:

This episode of the Event Tech Podcast explores the real-world performance of OpenAI's Operator, the agentic AI tool that promises to automate browser tasks on your behalf. Will shares his hands-on experience running Operator through its paces—from sending personalized Twitch messages to attempting LinkedIn research and even trying to buy festival tickets. Brandt and Will dissect what works, what doesn't, and why Operator currently feels less like a job-stealing super-agent and more like a well-meaning, but exasperating, digital intern. The conversation also touches on the future of agentic AI, the importance of user data, and the quirks of modern AI assistants.

Discussions Include:

• What “agentic AI” actually means and how Operator fits the bill
• Real-world Operator use cases: Twitch messaging, LinkedIn research, and online purchases
• The current limitations, quirks, and slowdowns of agentic AI tools
• Broader implications for the future of work and the data arms race among tech giants

Quotable Quotes (Should you choose to share):

"Once again, Will's taking the hit so that we don't have to, right? He's letting us know that it's not there yet." - Brandt Krueger

"I think we're potentially in this world where I do think agents are starting to do things that we never would have comprehended." - Will Curran

"It's like a really, really dumb intern. You show someone a task once and at least they get it, right? But this thing constantly forgets." - Will Curran

"The tasks that you’ve given it are very close to the examples we gave when we first started talking about operator and agentic AI. It doesn't sound like it can do that yet." - Brandt Krueger