I KEEP asking myself a simple question: what exactly are we training Filipinos to do with AI? There are a lot of programs and big numbers tossed around. Project AGAP.AI, an AI Academy, strategy roadmaps, training ten thousand “AI-ready” Filipinos. It sounds impressive. But when you dig into it, I wonder if we are really talking about meaningful skills or just teaching people how to use free AI tools and write a few good prompts.
Prompt engineering is useful. Anyone who has tried to get a chatbot to write a decent email knows there’s a knack to it. But is knowing how to phrase a request into ChatGPT or Bard the same as understanding the technology that is reshaping work and society? If our national response to AI stops at free tools and prompt tips, then I fear we might end up with surface-level skills that look good on paper but don’t translate to real advantage in the workplace.
I’ve talked to educators and workers. Many see AI classes as “show me how to use the app.” They expect step-by-step guidance on the latest shiny tool. But true AI training goes deeper than that. It should teach people why the tool behaves the way it does, where it fails, where it helps and how to combine human judgment with machine suggestions. That’s a different level altogether.
Look at what Singapore is doing for context. Singapore’s approach to AI education, which I’ve followed closely, aims not just to train users but to build creators and leaders in AI. They are investing in foundational understanding, ethics, data literacy and governance. They want engineers who can build models, officers who can set policy and professionals who can manage AI systems responsibly. Yes, they also have practical courses, but these sit on top of strong fundamentals.
If you compare that to a focus on prompt engineering with free versions of AI tools, the difference is clear. Prompt engineering is a tactical skill. It helps you get more from the tools available today. But without a strategic foundation — without understanding the math, the risks, the data behind the models — you are essentially teaching someone to drive without teaching them how the engine works or how to read the road signs.
This matters because AI is more than another productivity app. It is a set of technologies that can mimic parts of human thinking. In the 1990s, automation took over repetitive tasks. In the 2000s, the internet connected the world. In the 2010s, cloud computing made data and applications scalable. Those were tools people learned to use. AI feels different because it suggests, writes, analyzes and sometimes decides. That’s why people feel a mix of excitement and fear.
Our training programs need to reflect that difference. Sure, students and workers should learn how to interact with AI systems. They should know how to get outputs, assess them, and correct them. But they also need to understand limitations, biases and risks. They need to learn to protect data privacy, to think critically about AI suggestions, and to communicate outcomes to others who depend on their judgment.
Right now, the conversation around AI training in the Philippines gives me mixed signals. On one hand, there’s real government action. DepEd wants to train millions of students and teachers. Tesda’s Skills Passport promises free courses and certifications. DOST has invested in AI projects. On the other hand, the infrastructure challenges persist. Many schools still lack proper computer labs. Thousands need stable electricity. Internet access is uneven, especially in rural areas. Training without access is like building a library without books.
I believe Filipinos can compete in an AI-filled world. We’ve done it before with BPO services, with software development, with creative industries. We excel when we combine hard skills with ingenuity. But for that to happen with AI, training must be more than surface level. When people walk out of a workshop, they must leave with confidence and competence, not just a certificate that says they know how to use a free tool.
And there’s a bigger picture. AI will replace some jobs. But also, parts of jobs will change. What employers value most won’t be how well you can generate a prompt, but how well you can interpret AI results, make decisions, and add human insight. Asking, “can AI do this task?” is less important than asking, “what unique value do I bring that AI cannot replicate?”
If we want to close the global gap, we need training that grows thinkers, not just button-clickers.
We should teach people how to question outputs, how to spot errors, how to understand when the machine fails and why. Singapore understands this. They are preparing people not just to use AI, but to lead with it. That’s not a small distinction.
So, the fuss around AI training is justified, but only if we set the bar high. If we keep the debate stuck on prompt engineering and free tools, we will miss the bigger opportunity. AI training should not be about teaching people to chase the latest app. It needs to be about building capability that lasts beyond the next trend. If we do that, then the effort will be worth it. If we don’t, the fuss will look like noise, and we will have missed our moment.
The author is the founder and CEO of Hungry Workhorse, a digital, culture and customer experience transformation consulting firm.
