Equity

NEA's Tiffany Luck on AI IPOs, personal agents, and the ROI reckoning

35 min

Tiffany Luck, partner at NEA, discusses the magic moments in AI, particularly the rise of personal agents that can manage mental load. She shares insights on the upcoming AI IPOs of OpenAI and Anthropic, emphasizing that compute cost and unit economics will be key for public markets. The conversation also covers the ROI reckoning in enterprise AI, where startups are building tools to track token spend and optimize model usage, and the role of forward deployed engineers as a Trojan horse for adoption.

Summarized by Podsumo

🎙️ Ask about this episode

✨ Key Takeaways

1

Personal agents are almost ready for prime time, but need to overcome friction and security concerns to truly reduce mental load.
2

AI IPOs are exciting for the ecosystem, but investors will scrutinize compute costs and sustainability of growth.
3

ROI tracking is a hot startup opportunity, with companies like Ramp and Factory offering solutions for token spend and model routing.
4

Forward deployed engineers act as a Trojan horse, helping customers adopt AI while gathering feedback for product improvement.
5

The top 1% of AI-pilled firms spend $7,500 per employee per month on AI, but most companies are still in experimentation mode.

💬 Notable Quotes

"The magical moment will be when an agent can actually translate your mental load and take that away in a way where you actually believe it. — Tiffany Luck"
"I feel like we're only at like one to two percent of AI adoption, right? Across everyone. — Tiffany Luck"
"When a measure becomes a target, it's no longer works as a measure. — Tiffany Luck (quoting Goodhart's law)"