Return to site

Panel Discussion IV

broken image
broken image
broken image

AI Infrastructure to Support Innovation, in Global AI Race

Professor Konrad Young – CEO | College of Industry-Academia Innovation at NTUST

Sue Daley – Director | Technology & Innovation, techUK

Lawrence Lundy-Bryan Partner | Lunar Ventures

Josh Liu – Co-Founder & Director | Semi Ventures

This panel tackled the UK's ambition to become an AI maker rather than just a taker, emphasizing the importance of infrastructure investment, international collaboration, education reform, and strategic use of global supply chains. Taiwan’s role in data centers and semiconductor support was spotlighted throughout.

Key Highlights

1. UK’s AI Infrastructure and Global Ambition

  • The UK has committed £1 billion to double its compute power, aiming to become a top AI powerhouse.
  • Currently ranked #3 globally in AI, the UK seeks to own more of the infrastructure stack, not just import capabilities.
  • The AI Opportunities Action Plan targets training 7.5 million people in AI by 2030, with Chapter 1 focusing on compute infrastructure as a growth driver.
  • Emphasis on AI growth zones, supported by data center infrastructure and potential international partnerships.

2. Taiwan’s Key Role in AI Infrastructure

  • Taiwan is home to many of the top 7 global AI server and data center ODMs, offering deep hardware expertise.
  • Potential to partner with UK AI zones by exporting infrastructure, servers, or even supporting localized data centers.
  • Josh noted Taiwan's historical strength in building data centers and suggested it could help the UK replicate this success.
  • Calls for Imperial College–Taiwan internship exchanges to strengthen knowledge flow and bilateral innovation.

3. Cross-Country Supply Chain and Sovereignty

  • No nation can control the entire tech supply chain; strategic dependence on partners is unavoidable.
  • Countries like Germany (Pragmatic, Fractile) are exploring national champions—but success requires balancing protectionism with global integration.
  • Logistics and manufacturing decentralization were emphasized, referencing how Taiwan built up exports via Amsterdam hubs in the '90s.
  • Middle Eastern countries were cited as examples of new entrants into the semiconductor ecosystem, thanks to the modularity of the value chain.

4. Venture Capital & Hardware Investment Challenges

  • Hardware investment is capital-intensive, slow to scale, and requires massive funds, unlike software.
  • UK and Europe have niche advantages—like software layered with hardware differentiation—but scaling remains difficult.
  • The panel urged VCs and government funding arms to take a long-term view on hardware support.
  • Lawrence warned that unless the UK supports foundational tech, it risks falling further behind global competitors.

5. Education, Talent, and Job Creation

  • Mismatch between AI/quantum grads and actual jobs—graduates often face a lack of real-world career paths.
  • Call to link universities and employers more tightly through apprenticeships, skills task forces, or summer internship pipelines (e.g., with Taiwan)
  • Lawrence noted cultural issues: reluctance toward engineering careers, undervaluation of blue-collar tech jobs.
  • If jobs don’t exist, students should be encouraged to create them, fostering an entrepreneurial mindset.

6. Immigration & Global Talent Strategy

  • Strong consensus: Immigration policy must enable global AI/semiconductor talent mobility.
  • Reference to Taiwan’s historical semiconductor diaspora (e.g., brain drain to the U.S. during the golden age) as both a warning and a model.
  • The UK must compete for skilled immigrants, not just train its own, to stay in the innovation race.

Action Items

  • Enable partnerships between Taiwanese data center ODMs and UK AI zones for infrastructure development.
  • Launch internship programs between Imperial College and Taiwan to support AI ecosystem exchanges.
  • Push for immigration reform to help UK companies bring in AI/semiconductor talent efficiently.
  • Set up regional logistics and compute hubs (drawing lessons from Amsterdam’s former role) to support distributed AI infrastructure.
  • Support hardware VC funds and national champions, especially in hybrid tech areas.

Notable Quotes

  • “AI isn’t just a model—it’s compute, infrastructure, supply chain, and people.”
  • “No country can go full-stack alone—not even the US.”
  • “We don’t have enough jobs for AI grads. That means we need more founders.”