LCQ2: Enhancing employability of young people

TEXT START: The Government attaches great importance to and is committed to promoting youth development. In consultation with the Home and Youth Affairs Bureau (HYAB), the Education Bureau (EDB), and the Census and Statistics Department (C&SD), a consolidated reply to the Member's question is as follows:

DISSECTION

This document is a bureaucratic deflection dressed as policy substance. It answers three questions about youth unemployment and AI displacement with a catalog of existing programs, future studies, and statistical minimizations. Let me tear through it systematically.

THE CORE FALLACY

The entire reply is built on a lag-defense fantasy: that calibrated government responses — curriculum tweaks, retraining schemes, subsidy adjustments — can meaningfully address a structurally driven displacement problem.

The government explicitly acknowledges:
- 36,200 NEETs (6.3% of youth population)
- AI replacing entry-level positions
- Hong Kong's youth NEET rate (6-7%) persistently higher than Singapore and Japan even during economic booms

And then responds with: more STEAM targets, a fund for "innovative technology-in-education," a retraining board rebranding, and a manpower projections study due Q4 2026.

This is the institutional equivalent of adjusting deck chairs while the hull is breached. The framing assumes the problem is a skill gap that can be closed with better training. It is not. The problem is that AI eliminates the entry-level job category that has historically served as the entry point for productive economic participation.

HIDDEN ASSUMPTIONS SMUGGLED IN

1. Job replacement is temporary and offsetting. The reply cites studies that "AI would create new employment opportunities at the same time" as it replaces existing ones. This is the most dangerous myth in circulation. The DT framework is explicit: AI replacement does not create equivalent mass-participation employment at the same scale or timescale. The "new jobs" argument was plausible in the manufacturing automation wave (1950s-1980s). It is not plausible when AI performs cognitive labor faster, cheaper, and without burnout cycles.

2. NEETs are primarily "exploring career directions." The reply minimizes the structural component by attributing persistently elevated NEET rates to young people simply taking longer to figure things out. This is a comfort narrative for policymakers, not a diagnosis. If 6-7% of youth were structurally locked out even during labor shortages, the cause is not career indecision — it is a structural barrier that booms don't fix.

3. Skill-based training closes the gap. The entire Part 2 response assumes that adding AI content to vocational programs and funding tech-in-education will produce employable graduates. It will not. The logic fails on two grounds: (a) AI skills taught in 2025-2027 will be obsolete or automated by the time graduates enter the workforce; (b) the arms race in AI skills is won by AI itself — teaching humans to compete with AI at AI tasks is a losing proposition.

4. Policy review cycles match the speed of displacement. The government is conducting a "comprehensive analysis" with findings due in Q4 2026. AI displacement is not waiting for legislative timelines. By the time the manpower projections report is published, the landscape will have moved again. This is institutional process theater.

SOCIAL FUNCTION

This document performs transition management theater — it creates the appearance of active government response to maintain social stability and political legitimacy while the underlying structural displacement accelerates. It is not designed to solve the problem. It is designed to manage the political optics of the problem for a government that has no structural solution available within its ideological framework.

The NEET rate "mostly hovered around 6 to 7 percent even in years with robust economic growth" — this is the government's own data admitting that the problem is structural, not cyclical. And they still propose cyclical policy responses.

THE VERDICT

This is a death-delay memo from a system that cannot acknowledge the terminal diagnosis. The government has correctly identified the symptoms (AI displacement, elevated NEETs, persistent structural unemployment) and then prescribed the exact treatments that will not work (retraining, curriculum reform, subsidies, future studies).

Under the Discontinuity Thesis framework, every measure in this document addresses P2 (coordination impossibility) failures or P3 (productive participation collapse) symptoms through P2-compatible mechanisms (top-down retraining, institutional coordination) that cannot scale against P1-driven displacement. The Steam targets, the SSSDP priority for AI programs, the ERB upskilling scheme — all will be outrun by the very automation they are meant to prepare students for.

The government is essentially racing to teach people how to operate machines that will soon operate themselves.

The 36,200 NEETs are not a policy failure in the conventional sense. They are the leading indicator of the productive participation collapse that the DT predicts. The government's response — however well-intentioned — is structurally incapable of reversing it.

SURVIVAL ASSESSMENT FOR AFFECTED YOUTH

For the young people caught in this system:
- Conventional path (education → employment → career ladder): Increasingly conditional on becoming irreplaceable to a Sovereign, or indispensable within a niche that hasn't been automated yet.
- Lag-time reality: Some sectors (trade services, manual maintenance, regulated licensing) will retain human-only or human-primary roles longer.
- The government's promises: Fragile to terminal on the 1-5 year horizon. The $100 million fund for tech-in-education is not a survival plan — it is a hospice gift.

The government's own data is the obituary for the path it is advertising.