Eclipse reposted this
The rise of AI is accelerating the field of robotics, says Seth Winterroth, a partner at the venture firm Eclipse, which backs early-stage startups trying to solve hard engineering problems for physical industries. See other areas he's excited about in this week's edition of VC Wednesdays with Tanya Dua. 👇
Technology Editor at LinkedIn covering AI | Conference Moderator & Speaker | Columbia Journalism Grad | Ex-Business Insider
🚨 A partner at the VC firm Eclipse, Seth Winterroth invests in technology startups that improve the physical world, from autonomous vehicles to manufacturing automation. He joins us for VC Wednesdays. 🚨 ✒️ How would you describe your investment thesis? 80%-plus of global GDP comes from industries operating in the physical world. We saw an absence of any institutional capital being deployed into the earliest stages of company creation in those areas. The barriers preventing small teams from coming together to build full-stack solutions that can be deployed in physical industries had also dropped significantly. That was the genesis of Eclipse — to lead early-stage checks into founders trying to solve hard engineering problems to develop and deliver solutions into the industries that comprise the physical world. ✒️ What areas are you most interested in right now, and why? My focus is specifically on robotics and automation, and what we call embodied AI. Examples of that are Wayve, which is an end-to-end self-driving car company based in London, and Mytra, which focuses on automating manufacturing and the supply chain. I'm also interested in what I’d classify as high-performance compute. There's a big move to bring things back on-premise right now for a number of reasons, and I'm invested in a company called Oxide Computer Company that’s purpose-built for that across enterprise computing. ✒️ How has the rise of AI impacted robotics? Historically, automation solutions in the software layer were hand-engineered modules that you stitched together — from mapping to object detection and classification, and planning to control. AI has changed that, as you can train those modules on a significant amount of multimodal data and get it to perform a necessary task as good or better than a human. The same things that have enabled improvement in overall performance from GPT-2 to GPT-4 — data access, scale, compute and model architectures — are directly applicable to robotics as well. ✒️ Do you see defense tech funding continuing to accelerate? What ethical considerations do you look at while investing in the area? The post-Cold War era was a period of relative harmony, with the U.S. as the dominant leader of the world, certainly from a defense perspective, leading to the defense-industrial complex becoming consolidated and stagnated. But now, the U.S. has been caught flat-footed, and moreover, the nature of how we effect warfare has changed. That creates an opportunity. The best way to prevent warfare is through deterrence — it’s better to have it and not need it than need it and not have it. From an ethics perspective, I dearly love the values that we are fortunate enough to live by, and I want our society to be able to protect and defend them. To do that, it's really important that our defense-industrial base operates with the highest levels of capabilities and sophistication. #VCWednesdays #vc #venturecapital #startups #TechonLinkedIn