YouTube Summary · 30 Jun 2026
VIDEOSingapore's space strategy is not about launching rockets but about building the enabling deep tech (semiconductors, AI, quantum) that powers the global space economy. Deep tech is defined as frontier, nascent technology where the "rock is still being pushed uphill" — areas like fusion energy, quantum key distribution, and drug discovery. Talent development is the critical bottleneck, with programs like PowerX offering 70% subsidised on-the-job training to help mid-carere professionals pivot into deep tech.
Source: YouTube — Singapore Deep Tech & Space Strategy · Duration: 34:43 · Guest: Juliana Lim, Executive Director for People, SG Innovate
Build a skill that matches candidates to deep tech roles based on potential and mindset (not just keyword matching). Use natural language interviews to assess curiosity and learning ability — the video emphasised that the questions candidates ask reveal more than their answers.
✅ Actionable: Create a skill that conducts NL-based talent assessment interviews
Create a structured GBrain module for fusion, quantum key distribution, and space tech. Hermes can answer questions like "What is QKD?" or "How is fusion used in healthcare?" with curated summaries sourced from videos like this.
✅ Actionable: Seed GBrain pages for fusion healthcare, QKD, and SG deep tech ecosystem
Implement a PowerX-like agent that helps mid-career users identify transferable skills (e.g., TCM practitioner → biotech) and suggests subsidised training programs. The key insight: skills don't need to be directly relevant — lateral application of domain knowledge is valuable.
✅ Actionable: Build a career-pivot skill that maps transferable skills to deep tech roles
For startup founders, Hermes can provide personalised advice on non-monetary retention strategies (autonomy, purpose, innovation space) based on company size and team composition. The video's core insight: pay buys presence, not engagement.
✅ Actionable: Add retention strategy patterns to the startup advisory skill
An agent that scrapes government subsidy programs (like 70% training subsidies) and matches them to user profiles, automating the application process. Government intervention is justified when there's market failure — talent development is a clear case.
✅ Actionable: Build a subsidy-matching skill for SG government programs
For Hermes' own security, incorporate QKD concepts into agent-to-agent encryption recommendations. While practical QKD is hardware-dependent, the principle of quantum-safe cryptography should be flagged in our security review workflows.
✅ Actionable: Add quantum-safe cryptography check to security audit skill
Singapore's "no rockets" approach is pragmatic, but does it risk being a follower rather than a leader? If the global space economy reaches $1.8 trillion, Singapore's slice may be small without sovereign launch capability or satellite manufacturing. Being a component supplier in a rocket-free strategy means depending on others for the actual platform.
The video paints deep tech as inevitable, but many areas (fusion, quantum) are decades from commercial viability. Government subsidies may create a bubble of startups that cannot survive without continuous funding. The "rock being pushed uphill" metaphor cuts both ways — sometimes the rock rolls back down.
Calling AI "not deep tech" downplays the frontier research still needed (AGI safety, neuromorphic computing, edge AI). It may discourage investment in foundational AI research in Singapore. The fact that consumers use ChatGPT doesn't mean the underlying research is no longer frontier.
PowerX and Summation are small-scale. Can they truly address the talent shortage? The 70% subsidy is generous, but companies may still struggle to find enough candidates with the right mindset. The TCM-to-deep-tech example is inspiring but anecdotal — one swallow doesn't make a summer.
The claim that "pay isn't everything" is true, but in a high-cost city like Singapore, competitive salary is still a hygiene factor. Startups offering shares may not be attractive if the exit horizon is 10+ years and the startup mortality rate is 90%. Purpose doesn't pay the mortgage.
The video justifies intervention only for market failure, but deep tech subsidies may crowd out private investment and create dependency. A skeptic would ask: are we picking winners or propping up losers? The absence of a clear success metric for these programs is concerning.
The example of fusion for cancer treatment is essentially particle therapy (already done with cyclotrons). Calling it "fusion" may be misleading — it's using fusion byproducts, not fusion reactors. This conflation could undermine public trust when the actual technology is impressive enough on its own.