Uploaded July 2026 | Updated September 2026, 2 weeks ago
Paul Eremenko is co-founder and CEO of P-1 AI, the company building Archie, an AI mechanical and electrical engineer that works alongside human engineering teams. Jeff Immelt is a Venture Partner at NEA and former Chairman and CEO of GE, and he's joining P-1 AI's board after leading the firm's investment in the company's $50 million Series A.
Anne Dwane sits down with Paul and Jeff to unpack why engineering the physical world has lagged behind the AI boom in software, and how P-1 is closing that gap. Paul traces the idea back to his time at DARPA working on the F-35 program, where he saw firsthand how complexity, not procurement, was driving runaway costs, and explains how that thinking led to Archie's "semi-synthetic" training data approach for physical engineering domains. They dig into how Archie is built to sell as labor rather than software, why P-1 positions Archie as one AI engineer per team rather than a swarm, and how customer IP and data are ring-fenced under the company's deployment models. Jeff brings the perspective of a CEO who has advised dozens of companies on AI adoption, laying out his four-part framework for organizational readiness and the pitfalls of getting stuck in pilot purgatory. The conversation closes with a wide-ranging look at compute costs, hiring plans as P-1 scales into automotive and aerospace, and Paul's take on the technological singularity and what it means for how we build the physical world going forward.
Timestamps:
00:00 Cold open
00:32 Welcome to the Village Global Podcast
02:15 The origin of the name P-1 AI
03:06 Seeds of the company: Paul's time at DARPA and the F-35
07:37 What P-1 does with customers today
13:10 Why Jeff was drawn to P-1
14:59 Use cases and how ROI is measured
18:36 Enterprise adoption: workforce augmentation, not replacement
20:11 Jeff's 4-part framework for CEOs approaching AI
21:26 Avoiding corporate innovation theater
22:45 How customer IP and data are protected
25:44 Compute costs and Moore's Law
28:44 Hiring and building the team at P-1
30:43 Lessons from Radiology Partners on workforce enablement
32:19 Defining the technological singularity
34:49 What industrial OEM leaders should do to prepare
Paul Eremenko is co-founder and CEO of P-1 AI, the company building Archie, an AI mechanical and electrical engineer that works alongside human engineering teams. Jeff Immelt is a Venture Partner at NEA and former Chairman and CEO of GE, and he's joining P-1 AI's board after leading the firm's investment in the company's $50 million Series A.
Anne Dwane sits down with Paul and Jeff to unpack why engineering the physical world has lagged behind the AI boom in software, and how P-1 is closing that gap. Paul traces the idea back to his time at DARPA working on the F-35 program, where he saw firsthand how complexity, not procurement, was driving runaway costs, and explains how that thinking led to Archie's "semi-synthetic" training data approach for physical engineering domains. They dig into how Archie is built to sell as labor rather than software, why P-1 positions Archie as one AI engineer per team rather than a swarm, and how customer IP and data are ring-fenced under the company's deployment models. Jeff brings the perspective of a CEO who has advised dozens of companies on AI adoption, laying out his four-part framework for organizational readiness and the pitfalls of getting stuck in pilot purgatory. The conversation closes with a wide-ranging look at compute costs, hiring plans as P-1 scales into automotive and aerospace, and Paul's take on the technological singularity and what it means for how we build the physical world going forward.
Timestamps:
00:00 Cold open
00:32 Welcome to the Village Global Podcast
02:15 The origin of the name P-1 AI
03:06 Seeds of the company: Paul's time at DARPA and the F-35
07:37 What P-1 does with customers today
13:10 Why Jeff was drawn to P-1
14:59 Use cases and how ROI is measured
18:36 Enterprise adoption: workforce augmentation, not replacement
20:11 Jeff's 4-part framework for CEOs approaching AI
21:26 Avoiding corporate innovation theater
22:45 How customer IP and data are protected
25:44 Compute costs and Moore's Law
28:44 Hiring and building the team at P-1
30:43 Lessons from Radiology Partners on workforce enablement
32:19 Defining the technological singularity
34:49 What industrial OEM leaders should do to prepare
