The Call for Massive Upskilling: The Example of Hong Kong
The technological transition is accelerating worldwide, driving a profound transformation in the skills required in the job market. Recently, Hong Kong Financial Secretary Paul Chan announced the launch of the "AI for All" program, backed by a budget of 50 million Hong Kong dollars (about 6.4 million US dollars). This major initiative, managed by the Employees Retraining Board, aims to train more than 50,000 workers over two years across 200 targeted activities.
The stated goal of the Hong Kong authorities is clear: to prepare the workforce, particularly within small and medium-sized enterprises, for an unprecedented shift in the technological paradigm. According to a study cited by the Financial Secretary in his weekly blog, nearly 70% of Hong Kong residents already claim to have some familiarity or understanding of artificial intelligence tools. However, a significant gap remains between general consumer use and the productive integration of these technologies into business processes.
The Illusion of Coding for All and the Trap of "Vibe Coding"
Faced with this urgent need for upskilling, many institutions make the mistake of pushing workers toward intensive training in traditional programming languages (such as Python or JavaScript). Trying to turn every administrative assistant, teacher, or project manager into a software developer is not only unrealistic, but also inefficient. Acquiring advanced coding skills requires years of practice, a resource that organizations undergoing transition simply do not have.
Furthermore, the recent emergence of "vibe coding", a practice consisting of rapidly generating applications through simple text prompts without real structure or security auditing, raises significant concerns. The United Kingdom's National Cyber Security Centre (NCSC) recently warned organizations against the systemic risks of this approach. Without oversight, AI-generated code can introduce critical vulnerabilities, integrate outdated or compromised software libraries, and expose confidential data. A study published by the security platform Veracode confirms that more than 45% of applications developed using automated code assistants contain security flaws in their very first build. Upskilling, therefore, cannot rely on technical improvisation.
The Secure No-Code Alternative: Empowerment Through Simplification
Rather than forcing employees to learn to code, the solution lies in adopting no-code creation platforms designed to minimize technical complexity while ensuring a rigorous security framework. This philosophy of simplification and reducing the attack surface is precisely what drives the Quebec-designed ProductivIA application suite.
Within this ecosystem, the Fabrique application illustrates how artificial intelligence can be leveraged to support the citizen developer in a secure, structured way. Thanks to an entirely no-code interface, employees can describe the management tool their department needs in plain language. The Fabrique AI then generates a fully autonomous web application, written in standard code, but in a completely transparent and secure manner. Unlike unmonitored "vibe coding", the application is confined within a sealed sandbox and undergoes an automated security audit before deployment. End users never have to manipulate lines of code: they benefit from a custom, audited, and immediately functional tool.
To support this daily skills development, the ÉtudeIA application acts as a decentralized learning and continuous training companion. Leveraging RAG (Retrieval-Augmented Generation) technology, ÉtudeIA allows users to query an organization's document repository semantically. The AI's answers are not the result of statistical hallucinations, but are grounded in the company's actual documents (training manuals, internal policies, procedure guides) previously converted into mathematical representations called embeddings. Employees can thus master new business processes at their own pace, using a personalized tutor accessible in French.
Sovereign Infrastructure for a Trusted Transition
Workforce upskilling cannot come at the expense of data confidentiality. For government departments, school boards, or businesses subject to the strict requirements of Law 25 in Quebec, sending training queries containing strategic information to foreign servers poses a major legal risk.
The ProductivIA architecture addresses this concern by allowing all its applications to run on Matania, the locally hosted AI model provider in Quebec. Whether an employee uses the Courriel application to draft a message or the Fabrique application to design a business template, data never crosses national borders. Moreover, for organizations looking to upgrade their existing hardware at low cost, installing the sovereign operating system Boréal-OS breathes new life into computers declared obsolete by proprietary systems, while providing direct, secure access to the ProductivIA suite from a simple web browser.
Upskilling the workforce for AI should not be seen as a frantic race toward software engineering, but as an opportunity to simplify access to productivity tools. By replacing the complexity of code with the clarity of natural language interfaces, secure no-code puts control back in the hands of workers in a secure, ethical, and sovereign manner.