AI Salon: Building AI between Japan & Silicon Valley

Panel Discussion

Cultural Differences & Challenges

• Japanese founders often struggle with English communication and team building when expanding to Silicon Valley
• Risk-averse culture remains in large Japanese enterprises, but the startup ecosystem is evolving with government support embracing a "no failures means you do nothing" mentality
• Speed of iteration is significantly faster in Silicon Valley — one founder noted building a product in two weeks in the US that took six months in Japan
• Customer feedback in the US is more direct and honest, while Japanese customers have very high expectations for product perfection

Key Success Factors for Japanese Founders in Silicon Valley

• Communication confidence is critical — founders need either strong English skills or a partner who can communicate effectively on their behalf
• Team composition matters more than individual credentials — balance between technical expertise and business/GTM skills is essential
• Speed and iteration — ability to gather feedback quickly, implement changes rapidly, and ask direct questions about what's not working
• Traction is often required, though some VCs invest pre-traction if the team shows right balance of technical and business mindset

Investment Perspectives

Manufacturing & Physical AI (Omni Ventures)

• Global dynamism thesis: manufacturing and innovation will be distributed globally, not centralized
• Japanese companies often struggle to scale domestically first, then globally — different from typical Silicon Valley SaaS playbook
• Physical AI and manufacturing require 100% product quality, unlike software's "ship at 70%" approach — makes Japanese precision mentality valuable
• Smaller markets can serve as "stepping stone customers" to iterate products before scaling to larger US market
• Access to established manufacturing infrastructure in Japan is valuable for US founders looking to scale hardware production

AI & Robotics (99BC)

• Focus on "physical AI" — intersection of AI and the physical world, combining atoms and bits
• Four Ts framework for evaluation: Tech (technical moat), TAM (large addressable market), Team (product shipping experience), and Traction (customer conversations)
• Key moats: proprietary data access for software, intellectual property for hardware
• Hardware requires precision and extensive coordination across design validation, engineering validation, and mass production
• Five-layer cake framework (Energy, Chips, Infrastructure, Models, Applications) — different countries excel at different layers, requiring global collaboration

Japan-Specific Opportunities

• Demographic challenges creating demand for AI agents and automation to replace human labor in white-collar and service sectors
• Legacy manufacturers need AI solutions for compliance and documentation when selling into US markets
• Strong operational expertise and production scaling capabilities make Japan attractive for hardware startups
• University and research lab adoption is easier in Japan as starting point before enterprise

Geographic Considerations

• Geography matters significantly for model building (access to proprietary data) and manufacturing (access to local factories and infrastructure)
• Different regions have advantages in different parts of the AI stack — collaboration across countries is essential
• Markets large enough to support an auto industry are typically large enough to produce significant companies, but global-first mindset is still important

Founder Insights

Manas Kala (Veritas AI — Research Intelligence Platform)

• Had to unlearn perfectionism from Japan — US adoption happens faster if solving real pain point, even at 95% accuracy
• PhD background is double-edged: some VCs dislike "academic founders" for being slow, but helps with founder-market fit and credibility
• Not having Japanese co-founder makes it harder to work with Japanese institutions and universities
• In US, can access top professors and researchers without introductions, unlike in Japan
• Customer conversations are critical: team follows rule of talking to at least one customer every day and doing one reverse demo daily
• US campaign generated 100 leads with 30 wanting to talk; in Japan, much harder to get meetings with heads of labs or industry PIs

Cultural Learning & Adaptation

• American-to-Japan cultural transition appears easier than reverse direction
• Partnerships and distribution channels are critical for scaling — single partnership with major company can transform $99 Shopify sales into $30M orders within 12 months
• Velocity, feedback loops, and distribution are the key moats for startups

JETRO Support

• JETRO supports both Japanese founders expanding globally and non-Japanese founders wanting to enter Japanese market
• Multiple programs available for matchmaking Japanese companies with foreign talent
• Actively welcomes Silicon Valley investors to visit Japan and explore opportunities

Startup Presentations

Five companies from the JETRO program presented:

Event Format

• Panel discussion followed by networking session where attendees could meet the five startups at individual tables
• AI Salon format emphasizes human connection and honest feedback over formal pitching
• Startups encouraged to seek honest feedback, connections, and help beyond just funding