Hoy, kumusta? Welcome back to Kaisipan!
UNESCO just handed the Philippines our AI readiness report card. Yung grade? Aware but lacks proper execution.
After workshops across Luzon, Visayas, and Mindanao, UNESCO assessed five dimensions: legal, socio-cultural, economic, educational, and infrastructure. Main takeaway? We understand AI needs ethical governance, pero walang lead agency na nagco-coordinate ng lahat. DOST, DICT, and DTI all have separate initiatives. Walang single entity tying them together.
We're using the Data Privacy Act and Cybercrime Prevention Act as patchwork solutions, kasi wala pang AI-specific laws. Band-aid approach lang.
UNESCO's big recommendation? Build the National AI Research Fund under DOST and define AI ethics through a Filipino lens. Not copy-paste Western frameworks that assume infrastructure and context we don't have. Our history, culture, and values should shape our standards.
Here's the uncomfortable truth: Infrastructure remains the roadblock. Affordable broadband? Kulang. Digital literacy programs? Limited. How can we be AI-ready when basic internet access is still a privilege, not a right?
What this means for you: This report determines whether AI benefits everyone or remains concentrated in highly urbanized areas. Without infrastructure investment and policies reaching beyond NCR, AI tools won't reach farmers or startups outside Metro Manila. Poor execution means AI widens gaps instead of closing them. It means watching other countries surge ahead while we debate which agency should lead.
We are developing a National AI Strategy. We have awareness. We have private sector momentum. Pero execution requires sustained funding, political will, and real collaboration. Not just more reports and planning documents.
UNESCO gave us a solid roadmap.
The question is: will we actually build the infrastructure and policies, or will this join the stack of aspirational government documents gathering dust in some government office?
Kasi, right now, most people still use AI as a buzzword, not as a piece of technology that can augment lives.
Bes, Ano Yung…RAG?
RAG stands for Retrieval-Augmented Generation. Ito yung method kung saan ang AI model ay nagsesearch muna sa knowledge sources bago sumagot sa tanong mo. Hindi siya gumawa ng sagot out of thin air. Kinukuha niya yung info from trusted documents, tapos gagawa ng response based doon.
In short: RAG = Search + Answer. The model doesn't memorize everything. RAG feeds it the right information at the right time.
So what’s the point? Kasi kung walang RAG, ang AI ay gumagawa lang ng answer based on general training. It doesn't know your company's internal policies, your school's specific guidelines, or your latest project documentation.
So kailangan mong mag-upload ng documents like PDFs, spreadsheets, manuals. The system processes these files, breaks them into searchable chunks, and then retrieves relevant parts when you ask questions. That's how RAG gives you accurate, personalized answers instead of generic guesses.
Example: NotebookLM uses RAG. Kapag nag-upload ka ng documents, it searches through YOUR files first before answering. That's why it can cite exact sources and page numbers. Hindi general internet knowledge, kundi yung specific materials mo.
PhilRice Launches Palai, an AI Chatbot That Gives Farmers 24/7 Access to Rice Farming Expertise
PhilRice just solved a major problem for Filipino farmers: walang technical support after 5 p.m.
Launched during the 37th Ugnay Palay National Rice R4D Conference, Palai runs 24/7 through Facebook Messenger. Before this, PhilRice technical support operated 8 a.m. to 5 p.m., weekdays only.
Powered by large language models and retrieval-augmented generation, it pulls expert-validated answers from PhilRice's research database. The system understands complex, multi-part questions and learns from every interaction, refining its knowledge over time.
Lead Developer Mark Beltran explains: "Much like a child who grows more capable with experience, Palai learns from the questions it receives."
Why this matters for agriculture: Timing is everything in farming. Pest problems don't wait for office hours. Weather doesn't check calendars. Deputy Executive Director for Research Dr. Eduardo Jimmy P. Quilang emphasized this: "Even a one-day delay in addressing pest problems can lead to significant losses."
Palai democratizes access to PhilRice expertise. Location doesn't matter. If you have Messenger, you have access to science-based guidance whenever you need it.
The chatbot also automatically recognizes queries in any language or dialect, which is crucial in the Philippines, where language barriers often block farmers from getting timely help.
What's next: For 2026, PhilRice plans offline-capable mobile apps with voice services, plus a centralized database where LGUs can order knowledge products and track usage. They're exploring deployment in barangays and cooperative offices.
This is what practical AI implementation looks like. Identify the problem (limited extension services, time-sensitive needs), build the solution (a 24/7 chatbot), and deploy it where users already are (Messenger).
Mga Ibang Ganap
VERA Files launched SEEK, an AI chatbot that searches through 17 years of fact-checks and reports to help Filipinos combat misinformation and find verified context on personalities and events.
Packworks' AI tool boosted sari-sari store sales by 46% over 2 weeks, helping over 300 stores improve inventory management and demand planning with DOST-backed technology.
PLDT called for stronger AI investments in talent, governance, and infrastructure at the Digital Nation Summit, estimating AI could add 8-12% to GDP by 2030 if the country builds proper foundations.
SSS and PhilHealth launched AI-powered transparency portals to track government projects and spending, with DepEd, DOH, and DA portals expected to follow as DICT pushes for accountability.
Japan, US, EU, Canada, and Australia pledged cybersecurity and AI support to the Philippines during the Stratbase conference, with bilateral cyber dialogues and capacity-building programs lined up for 2026.
Mga Muni-muni
Leaks say ChatGPT is planning to introduce ads. The cat's out of the bag.
Android beta builds of the ChatGPT app now contain explicit references to an "ads feature"—strings like "search ad," "search ads carousel," and "bazaar content." Reports say OpenAI is testing ads initially inside the search-style experience and building an ad-buying platform to monetize free users by 2026.
Ads in ChatGPT will probably feel like "sponsored answers inside a conversation" rather than old-school banner ads. Free tier users will see sponsored suggestions or product cards, especially for queries like "best laptop for coding" or "where should I buy X."
Parang familiar, di ba? This is exactly what happened to Google Search. Remember when the top result was right at the top? Now you scroll through three to four sponsored results before seeing anything real. Ads didn't just add clutter but reshaped the entire incentive structure. Google now optimizes for ad revenue and "on-page time," not for quickly giving you the best answer.
Major queries now show multiple ad blocks before the first organic result. Relevant links? Buried. Non-paying sites? Invisible. Worse, ads look like real results and subtle "Ad" labels that are easy to miss, especially on mobile. People click, thinking it's the "best" result when it's really paid placement.
Ano impact nito sa end users? Answers may mix organic and sponsored content. Instead of a separate ad box, sponsored results could be blended as "recommendations" near the top of an answer. This blurs the line between neutral advice and paid placement unless clearly labeled.
Privacy concerns are real. Reports speculate that ads could be targeted using ChatGPT's memory or chat history. How much of your conversational data can be used for targeting? How transparent will controls be?
Research shows that intrusive ad formats quickly drive users away. If OpenAI keeps ads small, clearly marked, and context-relevant while leaving Plus/Team/Enterprise ad-free, the core "assistant" feel may survive. If not, people may start treating ChatGPT like an ad-heavy search engine and trust it less for neutral advice.
Prompt Tip!
Most people still think prompts are magic, like some perfect series of words that makes AI smarter. Spoiler: it doesn't work that way.
Ang kailangan mo is context, not complicated instructions.
Instead of telling AI how to do something step by step, tell it what you want to achieve and let it figure out the how. Research shows that goal-oriented prompts—where you describe the end state, not the process—significantly improve AI performance.
This insight comes from How to AI’s Ruben Hassid, who discussed the science behind what he calls "context engineering."
Here’s the formula:
[Primary Goal]: I need [output type] that [accomplishes what]
[Context]: This is for [audience/domain], where [what matters]
[Performance]: Success means [specific observable outcome], not [what failure looks like]
[Constraints]: Focus on [priority], avoid [anti-priority]
[Outcome]: The reader should [what they'll be able to do after]
Example: I need a strategy document that helps my team decide whether to adopt an AI tool for our workflow. This is for non-technical product managers who need to quickly understand trade-offs. Success means they can explain the three key decisions to leadership, not that they grasp the underlying architecture. Focus on practical implications and specific use cases. After reading, they should be able to explain why this tool is important for our particular situation.
Palaisipan
Is this AI or not?
That’s all for today!
Thank you for reading today’s newsletter!
Rate mo naman para I can improve my future issues. Ilagay mo na rin mga suggestions, opinions or violent reactions mo!
Palaisipan Answer: It is NOT AI. This image is from Robert Frank.