20 Questions on the Year AI Video Went Mainstream | A Conversation with Luma AI Product Manager Barkley: From Sora's Stunning Debut to a Crowded Field

Last week, we launched our "20 Questions" segment, and we're incredibly grateful for the support — it gave us the motivation to keep going. This week, we're using 20 questions to map the progress of another field: AI video models.

On February 15, 2024, Sora made its stunning debut, capturing the industry's full attention. Video models and video generation applications quickly became the focal point of the AI space. 2024 shaped up as a battle of many contenders: in Silicon Valley, there was Pika, Runway, and Google's DeepMind; domestically, players included Hailuo, Keling AI, Vidu, PixVerse, along with Tencent's Hunyuan and ByteDance's Doubao.

For this episode of Crossroads, we've invited Barkley, a product manager at Luma.ai, a leading Silicon Valley video model startup (and the sole PM on Luma's team, which has raised $160 million).

Through 20 questions, we'll explore the innovations and transformations in AI video models over the past year, and understand where the major players are headed. He'll also share his observations from the front lines as an industry participant — his take on the year since Sora's launch, his analysis of engineering and management capabilities versus algorithmic breakthroughs, and firsthand perspective on what people are discussing in Silicon Valley and how the PM role is evolving.

During our conversation, Barkley also told us about meeting Sam Altman at an afterparty and discussing whether vision is a necessary path to AGI. We hope this episode — packed with observations, analysis, and some industry gossip — proves useful.

P.S.: Barkley was formerly a overseas marketing intern for Koji at Tangdao. We're all typical "crossover" figures — from brand marketing and consumer products to tech, internet, and AI. This cross-disciplinary experience directly echoes the Crossroads theme: interdisciplinary thinking often yields unique perspectives and insights, and in this rapidly iterating AI era, such diverse backgrounds have become a source of differentiated competitive advantage. (In the timestamps, we haven't listed out the 20 questions directly; instead, we've woven them into the flow of conversation with Barkley. This not only presents the questions themselves, but also lets listeners experience the rich, dynamic exchange of the podcast format.)

01:50 Barkley's background: from TikTok's effects team to product manager at Luma.ai

03:11 One year after Sora's release, what paradigm innovations have emerged in video models?

05:16 What has Luma been working on these past few months?

06:43 What are the criteria for evaluating a video model?

08:27 Which video model is strongest? Currently, balancing motion amplitude, consistency, and aesthetics — who's doing better?

09:21 What are Pika, Runway, DeepMind, Hailuo, Keling AI, Vidu, and PixVerse each working on?

13:37 How does Silicon Valley view domestic Chinese AI video models?

15:12 DeepSeek's explosion has also drawn more attention to Chinese video models

15:47 Research or application — once you choose, then what?

18:52 If you choose the research-oriented path, how do you balance commercialization?

20:44 Video models and text models both follow Scaling Law in their development paths, but engineering training modes may differ

22:17 Is vision the necessary path to AGI? Barkley asked Sam Altman the same question at a party

24:46 What is the "world model" that Fei-Fei Li proposed?

26:48 The inspiration of "world models": not limited to generating video, possibly an anything-to-anything model

29:53 To continue breaking through, are engineering and management capabilities more valuable than algorithmic innovation?

31:58 What meaningful attempts have been made on the algorithmic breakthrough front?

33:25 When doing something no one has ever done, with no reference points, you must rely on an environment that boldly drives innovation

35:54 "Our CEO often asks candidates in interviews: for a problem that's never been solved before, how would you approach it?"

36:47 What's a video generation use case you've seen recently that made you do a double take?

38:27 Bold predictions: what happens in video generation in the short and long term?

41:19 What do people talk about during lunch at Silicon Valley companies?

43:35 The DeepSeek effect

45:46 Luma.ai vs. Runway — the clash between the two companies' CEOs on X

48:58 From being a effects product manager at TikTok to working on models at Luma.ai — how has the PM role changed?

52:39 What are product managers at other AI companies doing?

55:02 How have hiring requirements for product managers changed?

58:09 What do you do to keep learning in a rapidly changing industry?

59:28 What learning methods have you observed in others that are worth emulating?

1:02:03 For Chinese entrepreneurs in Silicon Valley, what new opportunities are different? How do you seize them?

Subscribe to the Crossroads podcast 🚦 We follow the industry transformations and new entrepreneurial opportunities brought by the new wave of AI technology. "Crossroads" was Steve Jobs' metaphor for Apple — standing at the intersection of technology and liberal arts, where great products are born. AI is transforming every industry. We seek out, interview, and bring together "active doers" of the AI era, exploring and embracing new changes and new possibilities alongside them.

👦🏻 Host Koji: Co-founder of The Fair / Tangdao. I believe technology, especially AI, will fundamentally transform society and empower humanity. Feel free to reach out to chat, exchange ideas, and connect on what's next. Koji on Jike, Koji's website

👧🏻 Host Ronghui: Works at a tech VC, former Silicon Valley correspondent for Yicai Weekly. Ronghui on Jike

Join the Crossroads membership community

☀️ First-hand AI news and insights

👫🏻 We encourage everyone to date / make friends / find future collaborators

🦀 Add our assistant on WeChat to join: Rwkfbcianvd, or scan the QR code below

🎄 This podcast is supported by The Fair's Sound Forest Podcast Project.