Philippine government agencies can use AI to leapfrog past their incompetence!

Analysis and Opinion

By Joe America

Corruption and incompetence are the foundations of the Philippine’s nagging third world condition. The rest of the world moves fast and rich to build modern economies, infrastructure, and citizen care.

Irineo B. R. Salazar reminds us that even the best of nations have to make fundamental changes to keep pace.

  • That even highly bureacratic Germany (something already noted by Mark Twain, as well as our long words) has managed to streamline paperwork to what is needed. And since the 1990s, there actually has been a push for service orientation in the bureacracy is something I would add. The days of the Kaiser when imperial officials sat way above the subject (not citizen) and yelled at them are mostly over, even as one still found holdovers of that mentality until the late 1980s.

Philippine government agencies and LGUs today remain a rats nest of paperwork, inefficiency, nonsense, and little dictators who seem to enjoy being rude to mask their incompetence.

This can be done away with fast. And it should be.

How?

Deploy AI administratively. Use AI rigorously throughout the agency or LGU.

Use AI to assess the current situation and operating environment. Look at strengths and weaknesses in operations.. Assess staffing conditions and needs. Examine work processes and outputs. Identify how to get more efficient. Develop 10, five, and one year plans

Develop project plans for all projects. Organize materials, labor, time frames. Identify barriers and how to overcome them.

Put citizen service as a priority and let AI recommend the scripts and the processes.

If you’ve read comments here at The Society of Honor, you can see AI being used to examine Philippine conditions and outline needed action. Karl Garcia is the best at drawing out AI thoroughness, updates, and ideas for action. Any agency could hire him, or someone like him, and positively leap forward.

Are other agencies in the way impeding progress? Let your AI master develop the coordinated solutions. Objectively.

If the guidance from AI seems not to make sense, order the machine to do better.

The AI systems are getting smarter and more powerful every day. They learn and produce, learn more and produce smarter.

Agencies and LGUs should get on this wave and surf it to the beach.

Stop with the incompetence.

Use AI as a general management tool. A powerful one.

_________________________

Cover photo from Triple i Consulting article “AI Technology in the Philippines: Transforming Business Practices for Growth“.

Comments
87 Responses to “Philippine government agencies can use AI to leapfrog past their incompetence!”
  1. Karl Garcia's avatar Karl Garcia says:

    This time it is on topic to use AI about the topic.

    Here I go

    You’ve touched on a cluster of very timely and interconnected topics: the Anti-Red Tape Authority (ARTA), the Philippine Competition Commission (PCC), the wider Philippine bureaucracy/line agencies, and how AI adoption might reshape their mandates, operations, and governance structures. Let me break this down into a structured analysis:

    —1. ARTA (Anti-Red Tape Authority) and AIMandate: ARTA’s role is to reduce bureaucratic inefficiencies, streamline procedures, and implement Republic Act 11032 (Ease of Doing Business and Efficient Government Service Delivery Act of 2018).AI Use Cases:Process Automation: AI-powered chatbots and virtual assistants can guide citizens through permits, licenses, and clearances, reducing face-to-face bottlenecks.Predictive Analytics: Identifying high-risk agencies/offices prone to red tape or corruption.Document Processing: Optical character recognition (OCR) + NLP for faster approvals and compliance checking.Monitoring Compliance: AI dashboards to track timelines for government transactions.Impact: A more transparent and efficient bureaucracy, but with risks of digital exclusion if agencies don’t ensure inclusivity (e.g., rural, elderly, non-digital citizens).—2. Philippine Competition Commission (PCC) and AIMandate: Ensures fair market competition, prevents monopolies/cartels, and investigates anti-competitive practices under the Philippine Competition Act (RA 10667).AI Use Cases:Market Surveillance: AI to monitor pricing patterns, detect collusion, or abnormal fluctuations in sectors (e.g., telecoms, energy, logistics).Merger Review: AI to simulate market concentration impacts of mergers and acquisitions.Algorithmic Collusion Oversight: Ironically, AI could help investigate AI-driven collusion (when companies’ pricing algorithms unintentionally—or intentionally—create anti-competitive behavior).Data Economy Oversight: As big tech and platform businesses dominate, PCC can use AI to audit algorithms for bias, consumer harm, or anti-competitive designs.Impact: Strengthens PCC’s enforcement in the digital economy, but also requires new expertise in algorithmic auditing and regulatory technology (RegTech).

    —3. Philippine Bureaucracy and Line AgenciesGeneral Opportunities:Public Service Delivery: AI can speed up services in agencies like LTO (Land Transportation Office), BIR (Bureau of Internal Revenue), DOLE, DFA, etc.Policy Analysis: AI-assisted modeling to simulate economic, social, and environmental impacts of proposed policies.Health & Agriculture: AI for disease surveillance (DOH), crop forecasting and fisheries management (DA, BFAR).Disaster Resilience: AI-powered risk mapping and climate modeling (NDRRMC, PAGASA, DENR).Law Enforcement & Justice: Predictive policing (PNP), case management (DOJ), legal research (courts).Challenges:Capacity Gaps: Agencies lack AI expertise; could rely heavily on private contractors.Fragmentation: Philippine bureaucracy is highly siloed; AI systems need interoperability and cross-agency coordination.Ethics & Privacy: Risks of surveillance misuse, bias in automated decision-making, or data privacy violations (intersecting with NPC – National Privacy Commission).Procurement Bottlenecks: Current procurement laws are not fully designed for agile AI adoption

    .—4. Cross-Cutting ImplicationsGovernance & Regulation: AI governance frameworks are not yet mature in the Philippines. Coordination among ARTA, PCC, NPC, and DICT (Department of Information and Communications Technology) is crucial.Public-Private Partnerships: AI adoption will likely depend on collaborations with tech companies, startups, and academe.Legal & Institutional Reform: Need to update laws to account for algorithmic decision-making, competition in digital markets, and digital rights.Inclusion & Trust: Citizens must trust AI in government—transparency and explainability are key.

    —✅ Strategic Opportunity:If aligned well, ARTA + PCC + DICT + NPC + line agencies could build a “whole-of-government AI strategy” that:Improves efficiency (ARTA),Safeguards competition in digital markets (PCC),Protects privacy and rights (NPC), andGuides digital transformation across all agencies (DICT).—Would you like me to draft a policy brief-style paper on this—something that could be used as a reference for governance reform and digital transformation proposals—or a matrix-style breakdown of agencies vs. possible AI applications (so it’s easier to visualize)?

  2. Joey Nguyen's avatar Joey Nguyen says:

    I’m not a big fan of general purpose LLM, otherwise marketed as “AI” with powers akin to science fiction artificial intelligence but without any real intelligence at all.

    My professional role is, to use the industry term, a “cat herder” of computer programmers and engineers. Popular culture has often made computer programmers to be some kind of all-knowing savants, a mantle many in the IT world eagerly adopted to gain social cred. But let’s be real, anyone who has had to manage IT resources knows that the amount of effort and cajoling required to focus personnel on major projects can be gargantuan when many line programmers are quite full of themselves and their own supposed smarts.

    So take “AI,” as the industry likes to call it, developed of companies flush with excess capital investment and little oversight, largely run by full-of-themselves programmers who have few guardrails, and think about what may be the resultant product. An “AI” is a computer program, and like any other computer program is only as good as the data it ingests and the ingestion routines for that data. Consider that most programmers and engineers are insulated, often awkward people. Okay a few may be real geniuses like Bill Gates, or even may be idiot savants, but most are not. They get ordered to write such and such code, and they do it. No different from a 1930s office secretary being ordered to type up a document on an IBM typewriter. Would such people be able to craft routines that can scratch the surface of interpreting human mental processes? Especially with little oversight? I’m afraid this is a moment of a small subset of people gazing at their own navels in self-satisfaction of their own mental superiority and passing it off as a major accomplishment.

    Anyway, it’s true that Philippines bureaucracy is a shit show. A lot of stuff makes no sense. There is a lot of redundancy. The Philippine bureaucratic process seems like a black box to be interpreted only by a self-appointed class of magicians. The nature of a black box is we don’t know what goes on inside with the inner workings, but we use it. So I’m not sure what is the utility of introducing a new black box of “AI” that is not created to understand the nuanced requirements of Philippine government. “AI” doesn’t have a track record of understanding English or American cultural norms that well, much less understand Tagalog, Cebuano, Ilokano, ALONG WITH understanding Filipino cultural norms and nuances. “AI,” does however, excels at reinforcing our own narcissistic beliefs of our own smug smartness reinforcing our preexisting biases. But that’s by design as bullshit generators that are programmed to emphasize user engagement.

    In order to solve the Philippines’ Byzantine bureaucratic processes, why not just do it the old fashioned way with reform? I’m not even a Filipino and I can identify at least 5 things most Filipinos I’ve met hate about the bureaucracy they need to deal with. Government reform, like better jobs, would be a popular item to campaign on. Make people’s lives easier, not harder. It just seems to me that introducing a new tool of “AI,” that the Philippines doesn’t even control, reduces accountability, makes stuff even more complicated and confusing, and now the bad bureaucrats (little dictators) have something else to blame if they are corrupt or incompetent. I always tell my programmers and engineers that keeping things simple goes a long way.

    • Karl Garcia's avatar Karl Garcia says:

      Let us say AI, or any computer project is rejected once more by Government employee group called courage, because of its luddite tendencies.

      How do they suggest

      Customs, DPWH , and other temptation prone agencies to straighten up? By self discipline?

      • Joey Nguyen's avatar Joey Nguyen says:

        Hmm. No one is born with self discipline. Come to think about it, every newborn only knows how to monopolize attention. Children often lie even about small things to get what they want or to get out of trouble. Discipline is a learned behavior that is instilled with model behavior from superiors paired with the threat of punishment. But if the superior is not righteous, it would be natural for those underneath to copy bad behavior, even more or when power allows room for hypocrisy to flourish.

        Good model behavior to be imitated is key.

        In the East Asian cultural sphere there is an expectation to moral righteousness rooted in Eastern philosophy that is displayed by older siblings, parents, teachers, government officials, leaders. The common saying across the East Asian cultural sphere is for a superior to “be a mirror” that those under them reflect. Yes, it existed in China as well before the Cultural Revolution destroyed Chinese culture. Those who get out of line are heavily penalized, sometimes by the legal process, but more often by cultural shunning.

        In the Western cultural sphere a similar expectation to moral righteousness is rooted in Christian morality. Well it gets one to the same end point.

        The corruption and graft issues with Customs and DPWH stem from historically bad leadership, too little punishment for wrongdoing, but most importantly too low salaries. One cannot expect to be a civil servant and get paid the big bucks. It’s supposed to be “service” after all. But there is a point that needs to be asked, “are those who enter civil service the best of servile leadership, or are they often the dregs of society?” In the Philippines, the answer is probably self-explanatory. A consequence of patronage.

        So what to do about it? Presidents need to appoint righteous leaders, like PNoy did, and like what Marcos Jr. half-did. Accountability means wrongdoers need to be made an example of, while those who are dedicated servants need to be lauded and rewarded appropriately with a salary that can support a middle class life. Do it often enough and corruption gets washed out. A stain on a piece of clothing often doesn’t get clean after one washing. Rinse and repeat. None of that requires the supposed capabilities of “AI.” If anything “AI” would introduce MORE unaccountability, more room for corruption, and more bad practices, all to the blamed on “the AI made me do it.”

        • three comments as I am on the way to the office today:

          1) I read a comment during the worst days of Duterte lamenting that the Philippines seemed to know neither East Asian shame nor Western guilt.

          2) supporting a middle class lifestyle even until retirement is very important. Civil servants in Germany get relatively little, but they have lifetime tenure as long as they “don’t steal silver spoons” as the saying here goes. I read a lot about ex-cops in the Philippines while an ex-cop in Germany is rare, sure they will do physically less grinding work before retiring but most retire as cops on government pension.

          3) there is the old Filipino joke (I was not surprised when both sonny and CV knew it) that the 11th and 12th commandments are “don’t get caught” and “if caught, don’t admit” and of course scapegoating those caught as a ritual is a corollary to that.

          There is now an online trend of mocking nepo kids of allegedly corrupt flood control contractors (with one normal businesswoman on X, Camille Co, getting annoyed that she is mistaken for a certain Claudine Co and bashed severely) but someone wise on X noted that the “wake and bake” bashing of Kitty Duterte was six months ago is already forgotten. What haffen vella is all I can say to all that.

          • Joey Nguyen's avatar Joey Nguyen says:

            Accountability also often means that there is a constant reminder that one has a lot more to lose for wrongdoing. Walking and living among DE’s during my visits has been a reward to me, but even with the intimate level I have, I’m not even sure if DE’s are ready if only penalty options are presented. Rewards should also be presented to encourage good behavior. Otherwise we have things like people scamming the local sari sari stores with sob-story based utang, then moving onto the next barangay when no tinderas trust them anymore. Chinoy business practices, which has a lot of elements of East Asian morality, has little acceptance for such behavior. Once a wrongdoer is caught word spreads and the offender is effectively locked out of any business dealings for at least a generation. They wouldn’t even be able to borrow 50 pesos, much less 1,000. There’s a reason why Bombays demand collateral for even small 5-6 loans. They get paid back almost every time. Constantly getting saved for bad decision making doesn’t let one learn anything. Help should be given if needed, but conditioned. Would that break the Philippines gift-giving cultural expectation? Maybe. Would it be better that the more negative aspects of gift-giving culture be broken? Probably. Other cultures have figured out ways to honor their old culture while moving forward. People need to graduate to being a responsible adult one day, and they should be given the education necessary to do so.

            As someone who is a beneficiary to the stability afforded by my father’s civil service, yes, a competent civil service is very important in building the middle class. Often families with civil servant parents in the West can go from poverty, to middle class, to graduating doctors and attorneys in just one generation. The stability of my father’s job after he left the military allowed my mother to open multiple businesses, which then allowed extra savings to go towards investments. Converting the 93% of Filipinos who are in poverty or desperate poverty via the civil service is important. But to do so the Philippines needs to upgrade schools and pay commensurate salaries to civil servants. If graduating teachers can earn more as a BPO agent… something is going wrong there.

            Personally I think there is a big problem with soft-theft in the Philippines. Often the victims are other family members who might be more “mabait” or “buotan.” They are practically treated as modern day slaves by their families. Of course we who read history know that the most common type of slave in pre-Spanish Philippines were extended relatives who were more poor, hah. People continue on these bad habits because they have no models to model off of. It doesn’t help that the common understanding is if one can get to some position of personal power, even as a tanod or captain, one can leverage it to make money. Here in the US our tanod equivalent (neighborhood watchmen) are all volunteers. As are block captains in larger cities.

            The biggest blocker to improvement in my opinion is that the educated and elite class live in their own little bubbles. They want to replicate their bubbles, while not wanting to give up any power, just so they can say that the Philippines is also a First World country. DE’s are already at the bottom, and raising them a few steps up the ladder is not as hard as it seems. Raising up someone starts with giving them easier access to the tools needed, the most important of which is education.

        • Karl Garcia's avatar Karl Garcia says:

          Thanks Joey for pondering.

        • JoeAm's avatar JoeAm says:

          I disagree that AI would foster lack of accountability. For one thing, you can’t go lower than bottom. For another, it is not political.

          • Joey Nguyen's avatar Joey Nguyen says:

            I work in the industry. The purported abilities AI is supposed to unlock is pure fantasy. Not sure how a tool that basically reorders data it ingested is more useful than a properly trained individual, of which there are plenty of Filipinos who are either trained already or are smart enough to be trained. MIT Press reported this week that 95%+ of AI pilots failed, evaporating billions of dollars in capital, perplexing many casual observers but is obvious to those who understand how the underlying technology works.

            What “AI” can be useful for is to *assist* in organizing data in order to speed up very defined programmatic workflows, provided the user knows what they are doing and can catch any LLM hallucinations. The problem with people seeing “AI” as a god-in-a-machine is that most users just don’t know how to catch errors. Even the AI stuff posted in comments here, I’ve observed routine erroneous information.

            I’ll maintain that what is needed is the start of reform, not introducing new variables that when those introductions inevitably fail will have blame piled on top of it as a convenient scapegoat rather than the policy makers and operators of such tools. Back when I gained weight due to over drinking in my 20s, the only solution was not some magic pills; the solution was hard work to mind my diet, reduce drinking alcohol, and regular exercise. All much harder in the 30s. The Philippines is like a pill culture, from PFDA approved vitamin pills marketed as a cure all to quack “panambal” sold by abularyos. Need to fix the root causes no matter how hard the lift, otherwise might as well start imbuing “orasyon” and smearing in spittle, hoping that fixes every ailment. There are no magic pills for easy fixes in life that replace hard work towards a goal.

            • JoeAm's avatar JoeAm says:

              There are no magic pills. Assistance in organizing data would be groundbreaking improvement over today’s chaos, lousy service, paperwork, and nonsense. Reform is an ideal that has no way of coming about, as we’ve seen tracking the nonsense for decades.

              • Joey Nguyen's avatar Joey Nguyen says:

                Let me reframe in another way based on your reply:

                If the existing process is full of chaos, provides lousy service, has too much paperwork and nonsense, would it not be more sensible to fix those issues? Not sure how an AI will help when the people who cause the problems in the first place are the ones pressing the buttons on the black box that no one understands. Doing it wrong faster is still doing it wrong.

                • JoeAm's avatar JoeAm says:

                  The agencies are inexperienced at modern management structures and order. They are a mix of political entitlement at the top, a layer of hardened veterans doing the same work as before, and hundreds or thousands of functionaries doing what they are told. There is no professional management structure and it is my belief that one person using AI to organize ideas and work would help the agency head and his deputies to see how to improve their output. I’m not suggesting that AI run the agency, only that it help reform how work is done.

                  • Joey Nguyen's avatar Joey Nguyen says:

                    Reform of the patronage system in the US started with citizen action clamoring for such as early as the 1860s. The Pendleton Act was passed in 1883 to create the Civil Service Commission to evaluate prospective civil servants based on merit rather than connection to patronage, creating the modern professional civil service. The Hatch Act of 1939 further instituted accountability for civil servants based on their merits rather than political connections. The process from civil agitation for reform to a professional civil service as an expectation took a generation in the US. In Eastern European and Baltic countries it took half a generation as they had the examples laid down by American and European reformed civil service that went before them.

                    I’m sorry to say, but this isn’t going to be an easy process for the Philippines. *IF* there is seriousness about it, at minimum civil service reform in the Philippines might take half a generation. I find it unfathomable to think that the Philippines would be able to do something completely new and transformational, disregarding all prior examples of successful actions done by others, and just go about it the “Filipino Way.” Alright, the Philippines has continually been doing that for decades, a century even. Where has that gotten the Philippines? Examples of what has been done successfully before by others are useful lessons to learn from. There is no substitute in life for hard work and patient consistency towards a firm goal.

                    • JoeAm's avatar JoeAm says:

                      Oh, the road is long and bumpy, arduously uphill and upwind, but no one knows what is around the corner. Lean forward, let gravity do it’s work, then quickly slide one foot forward to catch yourself. Then fall into the next step. Understand the reality and the possibilities and bring them together, one step at a time. No need to condemn the innocent or ignorant. They may have labored to build the road but they didn’t design it, and someone else is around the corner working on it.

            • Karl Garcia's avatar Karl Garcia says:

              I beg to differ, talented people can enhance their talent.

              Humility aside, if not for my proper framing of my questions, I would not get satisfactory results.

              GIGO and it’s opposte.

              It is not regurgitating

              • Joey Nguyen's avatar Joey Nguyen says:

                Sorry, I’ll have to disagree. If “AI” is used for basic information research, then that is not enhancing anything. That’s just a summary of dubious quality to be read once, perhaps twice, then forgotten tomorrow.

                • Karl Garcia's avatar Karl Garcia says:

                  Agree to disagree

                • Karl Garcia's avatar Karl Garcia says:

                  Every tool when not used stays in tyhe storage cabinet. It is how you use your tool. if your expertise before are resarching, AI can help a lot it is just like google. And you can use it alongside google. I completely question your downplaying of AI, even when you are from the industry.

                  • Joey Nguyen's avatar Joey Nguyen says:

                    Before graduating to more specialized tools that have caveats and drawbacks, it might be useful to learn how to use basic tools first.

                    • kasambahay's avatar kasambahay says:

                      I am probly speaking nonsense, here goes anyway:

                      I think we filipinos should create our own AI. I dont really like the idea of us being dependent on foreign AIs, with them controlling what we think, what to think, etc. which direction we should be thinking, sending us summat on a journey with destination unknown. we are put on the path which we have no control, cannot control, faced with AI’s unlimited power in equal measures of good and evil.

                      casualties are everywhere from celebs endorsing products they did not approve to the bullied child who can proverbially never leave the playground, the mob never dispersed, there is no privacy, no silence. AI records everything, publishes everything, it is sleepless, all seeing and often pitiless! we keep getting hit! the breaches and blows fall too often and we have yet to find ways to defend ourselves.

    • JoeAm's avatar JoeAm says:

      The intellectual and pragmatic might do not seem to allow reform. It is a rote culture I suppose, people waiting for good guidance that doesn’t come because it’s too hard so they go home. You’d think seeing a telephone pole in the middle of the new lane added to the highway would be recognized as a hazard but the reaction is “well, I see it” so everyone else can, too.

      AI provides the intellectual structure, not as flawed as real people trying to do new things, so can be the rote instruction that everyone wants.

      • Joey Nguyen's avatar Joey Nguyen says:

        A president is an advocate for new policy, as well as the one empowered to implement that policy. The president can install reform minded administrators, as PNoy did. Teodoro and Angara are two examples of reform minded administrators. A president can cajole the Congress into writing and passing common sense legislation, even accounting for the inevitable percentage of pork to keep congressmen and senators happy. If elected officials and appointed administrators are so disempowered, might as well not have a government at all!

        “AI” is not intellectual. I’ve worked in machine learning and language models which underpins “AI” off and on since the early 2000s, but it has existed since the 1970s. It is not surprising to me that the biggest proponents of the magical qualities of “AI” as marketed now are neither engineers or technical minded people. They are marketers. Language models cannot do anything other than what it was programmed to do, which is to output a different variation of what it ingested during machine learning. As I said, programmers just try to write code approximating what they were instructed to do, and when the business directive is fantastical to begin with, being based off of some charlatan executive’s childlike misunderstanding of how AI works in science fiction, the underlying problem is staring at us right there. AI only works best in carefully pre-defined use cases, provided the programmatic routine is tight. I doubt there will ever be a “general AI,” and the “AI” of today which is marketed as a general AI, is NOT a general AI. It is a word generator.

        Anyway, even suppose such an AI exists that can help streamline the bureaucratic process, if the real people using it are so incompetent they need to be assisted by AI, not sure how the result will be any different. The real people need to be trained. The organization may need reform, along with systemic reform. Reform which other countries have done either willingly or dragged into it kicking and screaming is the only real solution that exists today. I’d rather go for a tried and true solution that exists, over a *possible solution* that may very well not exist at all if ever, wasting money and time pursuing mirages.

        • JoeAm's avatar JoeAm says:

          Order is intellectual. AI can inspire reform by organizing it. A President is working an entrenched deep state of public and private powers that (a) keep him busy, (b) keep him distracted, and (c) block the right-of-way. President Marcos is talking the good game. Playing it is different.

          • JoeAm's avatar JoeAm says:

            I would add that you frequently call for action, but propose actions that simply can’t happen because the malaise is so deep. AI is different. It’s populist action, new, and not proven ineffective. It’s odd to find you discouraging the action.

            • Joey Nguyen's avatar Joey Nguyen says:

              I find it strange to characterize products made by techno-billionaires as “populist.”

              I am calling for action, which is sensible reform. It would be popular. Along with jobs. The executive has the authority and power to encourage, cajole, and if needed outright bully those who report to him in the executive departments into action. If need be the president must use the bully pulpit to browbeat Congress into assisting with legislation by putting a spotlight on incalcitrant inaction towards making the Filipino’s life better. If that seems too hard to even TRY to take the first step, then I’m afraid the Philippines is a lost cause.

              I do not believe in magic pills or magic solutions. Ignoring the problem and hoping some super intelligence (which doesn’t exist) to solve everything is not only lazy, but doesn’t fix the existing problem.

              • Karl Garcia's avatar Karl Garcia says:

                Many things are actionable.
                You may ask again, why no action?

                Example those allegedly snubbed by Jun Abaya later implemented by Dotr during the following admin.

                They no longer had to exert effort on due diligence because the due diligence has been done for them, all they need to fo is plug and play.
                Some others really gathered dust because it is hard to do unsolicited proposals from thin air.

                • Joey Nguyen's avatar Joey Nguyen says:

                  PNoy was trying to salvage what was left of EDSA and move the Philippines towards a better future. He had too much on his hands in 6 few years.

                  I recall this LRT “debacle” that benign0 on GRP used to incessantly trash PNoy with his nasty slur “BS” Aquino. I also recall that PNoy and Abaya were stymied by right of way issues from resistant LGUs that the LRT extension was supposed to pass through. Duterte steamrolled the same LGUs with veiled threats, which got the job done, but at the cost of consolidating more power towards the executive. Well I suppose Stalin also utilized unabridged state power to build grand projects like the Moscow Metro. Whereas in the Philippines the Dutertes continue harp on incessantly to this day about building out a small LRT extension, the groundwork of which was already largely completed by his predecessor.

              • JoeAm's avatar JoeAm says:

                By populist, I mean AI is trending and agency heads can use it as a tool to identify and justify reforms. People will be receptive to changes.

                • Joey Nguyen's avatar Joey Nguyen says:

                  If agency heads are in OVER their own heads, and cannot control the direction of the agency entrusted to them, perhaps they are not the right person for the job.

                  • JoeAm's avatar JoeAm says:

                    Some are, some aren’t. An agency is a deep state itself, a huge business lacking continuity of direction, budget, and political support, staffed by iconoclasts and newbies, and under changing leadership. There are shades of gray all over the place. I’m disinclined to say there is no opportunity to do better.

      • The stuff about telephone poles in the middle of new lanes reminds me of “not my job” satire pages on the Internet, where pictures like the one below are all over the place. Seriously that is more of a leadership issue. A lot of Filipino managers are smug while a lot of Filipino employees are scared to give feedback upwards, as the culture over there is often shoot the messenger.

        • kasambahay's avatar kasambahay says:

          ahem, that is funny! if there were filipinos around, they would have come in the middle of the night and burn the bush! and erase the double line, for april fools! but I am supposing the linemarker is AI operated and like a remote vacuum cleaner programmed to go around obstacles when cleaning, the linemarker did too, with nary a backward glance.

          and please dont think that filipino employee are so browbeaten they are so afraid to give feedback. they do give feedback their own way. you must have heard of bosses being shot by riding in tandem, their businesses set on fire! or bosses with bags of money to be deposited in the bank got robbed in broad daylight. or their businesses getting bad reviews in the internet. in this day and age, conflagration is fast getting to be the norm and the internet has freed the minions! anonymity in cyberspace has given them a voice. konektadong pinoy has lapsed into law (president marcos failed to sign it) and now, filipinos will be embracing modern technologies, both the bad and the good of it. but maybe more bad than good with scams proliferating and scam hubs hiding in plain sight, in techno hubs.

          as well, anomalous flood contractors are now turning against politicians, and being witnesses willing to swear oaths as to how far corrupt politicians can go, and how much money really did go into bent politicians’ pocket.

        • Karl Garcia's avatar Karl Garcia says:

          Many managers get fired by old social media screen shots, low evaluation scores. So I beg to differ on this no feed back feed back.

          • Karl Garcia's avatar Karl Garcia says:

            OMG, I just read kb’s comment.

          • my knowledge of that stuff is admittedly dated and social media is a double-edged sword that Filipinos use in different ways nowadays. For instance they are now gleefully calling out nepo babies whose parents got rich on flood control projects, revealing their expensive brand name handbags and all that they flaunted on social media. Reminds me of the Napoles kids, or even further back some high school classmates.

            I was also surprised that Bam Aquino and Kiko Pangilinan got into the Senate this year, just like I was extremely surprised when People Power happened in 1986, in a country described as 40 million cowards and an SOB by an American journalist. Still, I haven’t seen the big picture change that much, even as processes that take time historically look glacially slow in real life.

        • JoeAm's avatar JoeAm says:

          That’s about it. No one thinks “holy shit, someone (else) is going to get killed and it’s on me!” Accountability for others’ well-being is not in the cards, which I suppose is why democracy does not exactly click here. Democracy in the Philippines is a way to organize entitlements, not a way to achieve progress or take care of people.

  3. Karl Garcia's avatar Karl Garcia says:

    My late dad and his pma classmates filed a tax payer suit on the DPWH and the government which lead to procurement reforms slowly but surely.

    That is my lowdown, my knowledge and understanding is limited so I ask chatgpt yo turbo boost my understanding.

    Great angle. G.R. 167919 (Abaya v. Ebdane, 2007) is often remembered in a negative sense (because it allowed foreign-funded projects to skirt RA 9184 procurement rules). But if we flip the lens and think: what if the lessons and principles from this case were optimized?—then some bright sides emerge: 🌟 Bright Sides if Optimized 1. Clearer Procurement Rules for Foreign-Funded Projects

    • The ruling clarified that executive agreements and loan covenants (like JBIC rules) can override RA 9184.
    • If optimized: This could be used as a springboard to harmonize international lending guidelines with Philippine law, making procurement both compliant with global standards and protective of public interest.

    2. Strengthened Oversight in ODA-Funded Projects

    • The case highlighted the tension between national procurement rules and foreign donors’ conditions.
    • If optimized: The Philippines could design stronger monitoring and transparency mechanisms for ODA projects, ensuring that while foreign standards apply, Filipino taxpayers are still safeguarded.

    3. Empowerment of Taxpayer Standing

    • The Court recognized that taxpayers have standing to question questionable disbursements in foreign-assisted projects.
    • If optimized: This could empower more citizen suits, watchdog NGOs, and civil society interventions to prevent overpriced or anomalous deals early on.

    4. Push for Institutional Reforms

    • The decision exposed loopholes in RA 9184 and EO 40 regarding retroactivity and foreign loans.
    • If optimized: Legislators could amend procurement laws to close gaps, balance flexibility with accountability, and prevent future ZTE-like controversies.

    5. Capacity-Building for Philippine Contractors

    • The ruling indirectly showed how foreign contractors dominate big-ticket projects under ODA terms.
    • If optimized: It could be a wake-up call to support local construction industries, so that Filipino firms can compete with foreign bidders on quality, technology, and scale.

    6. Catalyst for Public Awareness

    • Though obscure to most citizens, the case foreshadowed the ZTE scandal. It taught that “small legal details” in contracts can have massive political and economic consequences.
    • If optimized: It could be used in law schools, governance training, and public education as a case study in vigilance, law, and development.

    In essence: The “bright side” of Abaya v. Ebdane lies in its ability to expose blind spots in governance and procurement, and if policymakers, institutions, and citizens act on those lessons, it could strengthen transparency, procurement integrity, and national capability in handling foreign-funded projects.

    Would you like me to draft a “what if” scenario (like: if the principles from this case were applied properly, the ZTE scandal might have played out differently), to make the bright side more concrete?

  4. Karl Garcia's avatar Karl Garcia says:

    I also asked why we are so vendor driven?

    Why not build capacity?

    That’s a sharp observation — the Philippine bureaucracy does tend to exhibit a vendor-driven dynamic, where agencies often become overly dependent on, or even captured by, suppliers and contractors. This problem is not unique to the Philippines, but local conditions amplify it. Here are some key points framing why this happens and its implications:—Why the Bureaucracy is Vendor-Driven1. Procurement DependenceAgencies often lack in-house technical capacity (e.g., ICT, infrastructure, specialized equipment). Instead of developing internal expertise, they rely heavily on external vendors who both supply the product and define the specifications.This creates a “specification trap,” where suppliers essentially write the terms that they themselves will later fulfill.2. Weak Institutional CapacityGovernment offices may not have the technical skills to evaluate competing products/services. Vendors exploit this by shaping bids around their proprietary solutions.This is especially common in ICT modernization projects, defense procurement, and medical equipment supply.3. Short-Termism in GovernanceBureaucrats and political leaders often want “quick wins” or projects that can be implemented within their term. Vendors present “turnkey solutions” that bypass the slower route of capacity-building.4. Procurement Law (RA 9184) ComplexitiesWhile RA 9184 aims to ensure transparency, its rigid bidding processes can paradoxically favor vendors that know how to navigate the system.“Lowest calculated bid” rules often reduce evaluation to cost rather than long-term value, encouraging supplier lock-in or underperforming projects.5. Corruption and KickbacksIn some cases, suppliers actively shape demand by offering incentives to decision-makers. This leads to overpricing, ghost deliveries, or purchase of unnecessary/unsuitable equipment.—Effects of Vendor-Driven BureaucracyLock-In and Monopoly: Agencies become dependent on one supplier, raising long-term costs (e.g., proprietary software requiring annual licensing fees).Misaligned Solutions: Vendors push products that may not fit the local context but maximize their profit.Wasted Public Funds: Projects fail or underperform because they were driven by supplier agendas rather than citizen needs.Erosion of Public Trust: People perceive government procurement as a venue for collusion and rent-seeking.—Examples in the Philippine ContextICT & E-Governance: Many government IT systems rely on proprietary platforms rather than open standards, making agencies dependent on foreign vendors.Defense & Security: Procurement of military hardware is often influenced by foreign suppliers rather than aligned with an integrated defense industrial strategy.Medical Supplies (COVID-19 PPE/Testing kits): Scandals showed how suppliers could dictate prices and quality, sometimes through politically connected middlemen.Infrastructure: Contractors often influence what gets built and how, with local governments reliant on “turnkey” contractor-managed projects.—Ways ForwardCapacity Building: Invest in in-house technical expertise within line agencies.Independent Procurement Offices: Strengthen institutions like the Government Procurement Policy Board (GPPB) with more technical oversight powers.Open Standards & Open Source: Reduce lock-in risks, especially in ICT.Lifecycle Costing: Shift from “lowest bid” to value-for-money frameworks.Stronger Transparency & Oversight: Enhance use of e-procurement platforms with public monitoring (e.g., PhilGEPS transparency).—👉 Would you like me to frame this into a policy brief style (like the one you asked for with ARTA/AI), highlighting problem, evidence, and recommendations? That way, it could be a sharper piece for decision-makers or commentary.

    • Karl Garcia's avatar Karl Garcia says:

      Case in point: Frigates aquisition.

      You’re pointing to a core structural problem in Philippine defense procurement: acquisitions like the cutter deal with Hyundai Heavy Industries (HHI) and the compatibility issues with combat management systems (CMS) aren’t isolated technical hitches, but symptoms of deeper institutional weaknesses. Let me break this down:—1. Case Example: Cutter Acquisition with Hyundai Heavy IndustriesThe Philippines acquired Jose Rizal-class frigates from HHI. While the ships themselves were delivered, the combat management system (CMS) integration sparked controversy.The original requirement reportedly preferred Thales TACTICOS, but HHI instead pushed Hanwha Naval Shield, creating disputes about compliance with specifications.This highlights the vendor-driven nature of procurement: instead of the AFP setting non-negotiable standards, suppliers often dictate what is “available” or “affordable.”—2. Systemic Issues in Defense ProcurementVendor-Driven DealsForeign suppliers often dominate negotiations due to local lack of capacity in shipbuilding, electronics, and weapons systems.This means procurement tends to be “off-the-shelf packages” bundled by foreign contractors, rather than modular, interoperable systems designed around Philippine doctrine.Weak In-House Technical CapacityThe DND and PN have limited systems integration expertise. Without in-house naval architects, CMS specialists, or electronic warfare engineers, they cannot fully assess whether a vendor’s offer meets operational requirements.Reliance on foreign consultants and contractors creates a cycle of dependency.Procurement FragmentationDifferent projects are pursued without a strong long-term force structure vision.Ships, planes, and radar systems may be modern individually, but interoperability suffers (different CMS, incompatible datalinks, no shared doctrine).Budget and Political PressuresTenders are often awarded not only on performance but on political alignment, financing, or “lowest bidder” rules.This prioritizes immediate affordability over lifecycle cost, maintainability, and combat effectiveness.—3. ConsequencesOperational Gaps: Navy ends up with modern hulls but limited sensor/shooter integration.Higher Lifecycle Costs: Dependence on vendor-specific spare parts and service contracts.Strategic Vulnerability: Without indigenous capacity, the Philippines is locked into foreign supply chains, which can be cut or leveraged geopolitically.—4. Path ForwardTo move away from a purely vendor-driven, “lowest-cost” mentality, the Philippines could:Develop In-House Expertise: Invest in training naval engineers, electronics specialists, and CMS integrators under NEDA/DTI-DND coordinated programs.Independent Technical Authority: Establish a defense acquisition & evaluation body insulated from supplier lobbying, capable of validating CMS and subsystem integration.Modular Procurement Approach: Require open-architecture CMS standards so that sensors and weapons from different suppliers can be integrated in future.Strategic Industrial Partnerships: Instead of turnkey projects, push for tech transfer and local assembly, e.g., strengthening Subic shipyards or other local facilities.Whole-of-Government Oversight: Broaden involvement beyond AFP—tie procurement to long-term industrial strategy (like how Indonesia develops PT PAL with Korean and Dutch help).—👉 This raises a larger question: do you want me to frame this as a policy brief (diagnosis + solutions), or as a case study analysis of the Hyundai/Philippine Navy cutter acquisition? Both can work, depending if you want to target policy reform or historical critique.

      • Joey Nguyen's avatar Joey Nguyen says:

        The Philippines procurement process across all government sectors is vendor driven because there is no consistent government-led industrial policy. Even during PNoy’s time there was none, as PNoy was too busy plugging the leaky national bangka so that at least the Philippines can continue a forward course.

        For example, to make naval ships one must first make commercial ships. To make commercial ships one needs steel and shipyards. To make steel one needs to build steel factories. To build steel factories one needs to build blast furnaces to fire the iron stock. To make iron stock one needs to build ore smelters. To get iron ore one needs to either import iron ore or mine it. All the above needs an energy source at each step, usually either coal or natural gas in modern processes. To get coal or natural gas coal mines need to be expanded, or natural gas extraction needs to be pursued. It’s not as simple as technology transfer. Filipinos may need to work under foreign investing companies while gaining experience, until a native Philippines company buys out the operation or a native company is started.

        It might be helpful to start with the first step instead of thinking of the last step. Focusing on the last step in an effort to jump to greatness is the greatest folly of the Philippines. One needs to learn how to walk before running. One needs to learn how to crawl before walking. The Philippines is like a country that is in the crawling stage but insists with little effort and too little experience, long distance running can be accomplished. Let’s get crawling right first.

  5. https://x.com/BayaniChain/status/1960880082366685642 Bam Aquino has interesting ideas:

    “You can cook the books, but not the chain.

    Sen.
    @bamaquino
    is calling for the national budget to be recorded on blockchain where every peso spent is traceable and every project has a digital receipt.

    In a country where ghost projects and missing funds make headlines, this could mean real accountability.”

    • Karl Garcia's avatar Karl Garcia says:

      Blockchain already IS being used in National Budgeting.

      Absolutely — the Philippine government has already started using blockchain technology in managing parts of its national budget, specifically by making critical budget documents publicly verifiable and immutable. Here’s what’s happening:

      Current Use of Blockchain in the National Budget

      The Department of Budget and Management (DBM) launched Project Marissa, a groundbreaking initiative that integrates blockchain into its Action Document Releasing System (ADRS). This ensures security and tamper-resistance for vital budget documents, including Special Allotment Release Orders (SAROs) and Notices of Cash Allocation (NCAs) .

      Through the platform hosted at blockchain.dbm.gov.ph, the public can access immutable versions of these budget documents via the Polygon blockchain, leveraged through Bayanichain’s blockchain-as-a-service (BaaS) platform and the Prismo protocol .

      This publicly accessible system went live on July 30, 2025, marking a significant leap toward transparency in public finance .

      Future Vision: Entire National Budget On-Chain?

      Senator Bam Aquino has proposed an ambitious initiative: to put the entire Philippine national budget on-chain, making every peso spent traceable in real time .

      If approved and implemented, this would build on the existing blockchain infrastructure and create an unprecedented level of accountability and transparency in government spending — potentially making the Philippines the first country to fully operationalize a blockchain-based national budget .

      Overview Table

      Aspect Current Status Future Proposal

      Blockchain Usage Yes — used for SAROs and NCAs via Project Marissa Proposed — extend to entire national budget
      Public Access Yes — via blockchain.dbm.gov.ph Would allow full real-time traceability of all government spending
      Underlying Technology Polygon blockchain, Bayanichain’s Lumen BaaS, Prismo Protocol Expand same tech to full budget
      Launch Date July 30, 2025 Proposal surfaced August 2025
      Significance Enhances transparency and document integrity Could vastly deepen public trust and accountability

      Final Take

      Yes — blockchain is already being used in the Philippine government’s budget system, but only for specific documents (SAROs and NCAs) so far. The broader adoption of blockchain for the entire national budget remains aspirational and currently under legislative consideration.

      Would you like more details on how Project Marissa works technically, or updates on the legislative process for the proposed on-chain budget?

  6. Karl Garcia's avatar Karl Garcia says:

    KB is right,chatgpt already tells us what we already know

    If I have no idea about what I am asking and what to ask. CHATGPT AND the rest can not help me or any one else.

    • JoeAm's avatar JoeAm says:

      Right. Prompt writing can start blind or broad though, then walk into greater layers of “known” material that AI can find in seconds that would take us a lifetime to ferret out. AI is a knowledge probing tool. How that can be bad is beyond me. The big tech companies are investing billions. I figure they have pretty good understandings of it. Stock investors are into it big time as well judging from tech trends.

      • Karl Garcia's avatar Karl Garcia says:

        Correct.

        • JoeAm's avatar JoeAm says:

          I have to register an e-mail to read the article and I don’t do that. I’ve wondered for a couple of years now at tech company valuations, most of it seemingly keyed to AI investments. I’ll have to examine Microsofts balance sheet and income statement. Presumably investors are doing than and aren’t seeing bubble, they’re seeing revenue. Housing forms bubbles because lenders overextend, house values inflate, then the economy crashes. Maybe stock markets are indeed over-extended for revenue potential. And maybe there will be a Trump crash next year that reveals an AI bubble. But the point is the scope, the scale, the players. It is a big deal, a real deal. The tools are available, learning learning learning. I’m no luddite.

          • Joey Nguyen's avatar Joey Nguyen says:

            AI stocks have insane P/E ratios while not having profitable products. The market valuation is consistent with meme stock status from institutional investors who are flush with excess capital they are essentially gambling with, rather than investing in product R&D that have a higher likelihood of yielding tangibles. Already OpenAI has moved the goalposts from “AI is on the level of human intelligence” to “Artificial General Intelligence (AGI) will unlock human level intelligence,” to “AGI isn’t a ‘big deal’ or goal.” Sam Altman himself has said recently that AI is in a bubble and cautioned against investing recklessly. I seem to recall that not too long ago FTX’s Sam Bankman-Fried, the poster boy of crypto, was heralded as a poster boy of cryptocurrency being the new great thing. SBF is now sitting in prison for massive securities fraud, his company imploded. “AI” is a chatbot that regurgitates information it ingests, cannot generate any new information, and a lot of the emotional tricks used by AI to make it seem like it has that capability are in fact HARD CODED fixes.

            SeekingAlpha is one of the most trusted tech industry analytics firms. I pasted the analysis article as follows:

            ***

            The AI Bubble Is Leaking: Prepare For A Major Market Reversal

            Aug. 22, 2025 4:48 PM ETSPDR® S&P 500® ETF (SPY), DIA, QQQ, IVV, VOO, IWM, SPX, SP500, NDX, DJINVDA, PLTR, META, SPY, VNQ, UTF, O, RYN, MAA, EPD, ET, BIP, BIPC, OKE, BIP.UN:CA, BIPC:CA, DIA, QQQ, IVV, VOO, IWM, SPX, SP500, NDX, DJI

            Samuel Smith
            Investing Group

            5
            Share

            Save

            (9min)
            Comments
            (353)
            Summary

            The AI bubble appears to be teetering on the brink of bursting.
            The stock market is flashing several warning lights.
            We share how we are positioning our portfolio right now, including some of our top picks of the moment.
            Looking for a portfolio of ideas like this one? Members of High Yield Investor get exclusive access to our subscriber-only portfolios. Learn More »
            Contrarian Investing. Opposite trend.
            Olivier Le Moal/iStock via Getty Images
            Ever since ChatGPT took the world by storm and NVIDIA (NVDA) delivered blowout guidance in the early months of 2023, the AI trade has dominated the markets. Whether it be leading AI players like NVDA and Palantir (PLTR) or some of the traditional megacap tech powerhouses like Meta (META), AI-related stocks have far and away crushed the broader markets over that period of time.

            Chart
            Data by YCharts
            While AI is truly an exciting and revolutionary technology that I expect to deliver incredible value to the global economy in the coming decades, cautions being raised by the likes of OpenAI CEO Sam Altman—that AI is in a bubble and that investors are getting overly excited—should give investors pause about what price they are paying for exposure to future earnings power from AI.

            Between these indicators and very frothy market valuations, I believe that the AI bubble is starting to leak and that prudent investors should position their portfolios for a major market reversal. In this article, we will detail why and share how we are positioning our portfolio accordingly.

            Growing AI And Market Concerns

            While analysts have been growing increasingly bullish on the company, raising their growth estimates even as its stock price has pulled back meaningfully recently, PLTR is perhaps the poster child for why AI is in a major bubble. It currently trades at a whopping 210 times forward earnings and 518 times trailing earnings. While this may not seem like an outrageously high valuation given that analysts are expecting it to deliver 35.7% CAGR in earnings per share through 2029, even if you assume that it achieves that growth rate at its current stock price alone, its 2029 P/E ratio would still be an incredible 73x.

            tikr.com
            tikr.com
            This means that PLTR is going to have to sustain a very high earnings per share growth rate for many years into the future to justify its current valuation and deliver a reasonable rate of return to shareholders buying at today’s price. As a result, I think that it could easily lose another 50% or even more of its value, based on its current very rosy growth estimates from analysts, and still be considered richly priced.

            On top of that, MIT recently released a study showing that a whopping 95% of companies pursuing generative AI initiatives are seeing little to no financial returns on investment. While there are many reasons for this—including challenges with integrating generative AI into existing workflows, human error, and other inefficiencies—it still raises serious questions about AI’s ability to deliver significant value for companies in the near term.

            As such, this study further accentuates concerns that AI may be in a bubble, with investors getting overly excited about AI’s capability to create real-world value on a large scale in the near future. It also raises major doubts about Palantir’s valuation, as just because companies are taking them on and the U.S. government is giving them increased revenue does not necessarily mean that they will be able to deliver commensurate value in exchange for that money. This could challenge the durability of those contracts and, in particular, the durability of their growth runway.

            Even if they are delivering value for that money, if the broader perception shifts such that AI is not as value-add as previously thought, companies may become more hesitant to pay the large sums required to bring on Palantir and similar firms to try to enhance their operations. Additionally, if AI is not as value-add as previously thought, it likely means that major technical breakthroughs are still needed before it can deliver on its promise.

            These include substantial improvements in agentic AI, which could eventually displace significant white-collar jobs once LLMs are able to operate autonomously on computers to perform a wide range of tasks that currently require office employees. Another major breakthrough that will likely be required is significant advances in autonomous robotics, which would enable AI to perform real-world tasks ranging from assembly line work to restaurant operations, logistics warehouses, and even agricultural labor. Robots could also eventually take on many tasks currently performed in the construction industry and some of the trades, which in turn would devastate large swaths of blue-collar work. However, the technologies needed to truly displace humans in both of these domains still seem quite far away.

            As a result, the question now becomes: how long are investors willing to wait to see true bottom-line profits from AI? Will they be willing to wait patiently while these companies burn tremendous amounts of cash engaging in R&D and infrastructure build-out, or will they begin to grow impatient and demand profits to begin hitting the bottom line? If the latter becomes the reality, then we could see a major pullback in the valuations of AI stocks ranging from PLTR to the leading megacap tech stocks.

            How I Am Investing Right Now

            With those concerns in mind, what does this mean for how we are positioning our portfolio? When you consider the bubble-like conditions and the risks for a short-term pullback in the AI space, and then extrapolate to the broader S&P 500 (NYSEARCA:SPY) where leading valuation models also seem to indicate that the market is highly overvalued, and couple that with numerous geopolitical and macroeconomic risks—such as the growing risk of conflict in the Far East, ongoing tensions and proxy wars between Israel and Iran in the Middle East, uncertainties regarding the outcome of the Russia-Ukraine war, continued tariff risks, runaway deficits in the United States and elsewhere, and the ever-present risk of inflation re-accelerating and what impact that could have on long-term interest rates in the U.S., especially when paired with persistent deficit spending—there are many reasons to be concerned about investing in broader equity markets today.

            In fact, all it would take is an AI bubble bursting, and the S&P 500 would likely plunge, given that tech stocks make up about a third of its underlying portfolio. As a result, we are steering clear of broader index funds as well as megacap tech stocks. Instead, we are looking at areas of the market where there are strong fundamentals, recession resistance, and very attractive valuations. We think these investments should do well regardless of what happens in the macroeconomic setting or how the AI saga plays out.

            In particular, we like REITs (VNQ) and infrastructure (UTF) right now, with blue-chip opportunities like Realty Income (O), Rayonier (RYN), Mid-America Apartment Communities (MAA), Enterprise Products Partners (EPD), Energy Transfer (ET), Brookfield Infrastructure Partners (BIP)(BIPC), and ONEOK (OKE), all offering highly compelling dividend yields and valuations backed by strong balance sheets and durable and defensive business models.

            Investor Takeaway

            The AI boom has generated tremendous wealth over the past several years, and we have admittedly missed out on much of it due to our focus on income-generating securities. We have nonetheless managed to outperform the market through our opportunistic capital recycling program. We believe that our portfolio is positioned for particularly strong performance moving forward, given our focus on AI-agnostic, blue-chip, deeply undervalued stocks, and it should benefit from a major market rotation out of AI into more profitable names as investors grow more concerned about the profitability of AI in the near term.

            On top of that, with bubble-like market valuations and numerous macro downside catalysts brewing, I think that right now is a prudent time to take a more defensive posture in one’s portfolio. This is the approach we are taking at High Yield Investor.

            • JoeAm's avatar JoeAm says:

              Thanks for the article that makes good reading, if a tad nerve-wracking. I do find comfort in knowing that houses are still there after numerous crashes. Banks, not so many, lol.

            • kasambahay's avatar kasambahay says:

              my feelers are bristling! you mention nvidia and the people are getting summat antsy with nvidia, there is apparent personality clashes with nvidia ceos, farqhuar and cannon brookes. the two used to be co-founders, now going their separate way. the company they both founded maybe pulled in different direction, with farqhuar resigning while cannon brokes stay to become sole ceo. investors are watching, but nvidia seems to have very good pr machinery and all seem well apparently: on the outside, the inside is yet to be seen. the next general meeting should be interesting.

      • I see AI as one step beyond search engines, and I remember the days when they were really BAD, and suddenly my brother told me to try Google for the first time. AI will get better but like search engines will rely on the quality of data available. As for the bubble, there was a dot com bubble in the late 90s, a bit of a weak period and a decade later smartphones and Facebook made the stuff only dreamt of in the late 90s reality.

        My brother BTW was for a while sales (or was it marketing) head of a short-lived Munich startup that tried to create something like an iPad in the LATE 1990s. Time simply wasn’t ripe for it, neither today’s batteries nor the Wifi or 5G of today existed. The contraption was almost as heavy as a laptop of today and extremely flaky. What I think is that the bubble will burst, but after that AI will continue to penetrate more and more mainstream IT, quietly – it already is.

  7. Joey Nguyen's avatar Joey Nguyen says:

    Taking a page from JP, I must take a step back from commenting on Philippines issues for a while.

    Tomorrow I’ll be flying out to Paris to visit family I haven’t seen in a while, then onward to Rzeszów, crossing the border into Lviv to meet up with an aid group I work with supporting the Ukrainian effort at the front delivering equipment and trucks. Once I make my way to the east and south of Ukraine near the front line, connectivity will be spotty for days at a time. Also an opportunity to do lay ministry with the Redemptorist Order.

    I will be in the Philippines in November to deploy donated solar irrigation equipment our Catholic charity has collected funds for, and I have helped to design, on calamansi farms in Zamboanga. So I’ll probably focus on this and the former. My Philippines activities will be in areas I can directly affect.

    Then back to the US where I will be focused on democracy issues because if the US ends up being permanently screwed, the rest of the world will likely be screwed too pinched between the Russians and Chinese who seek to dominate Europe and Asia.

    Cheers to all. And take care.

    • JoeAm's avatar JoeAm says:

      Have a terrific time, Joey. It sounds like you will. Riches in ways other than money.

    • Karl Garcia's avatar Karl Garcia says:

      Enjoy, Live long and prosper! Thanks for your time here.

    • All the best. So you shall be on three fronts of the present global struggle: home front, European front and Asian front. I will continue to focus on the my home front (Europe) with an occasional glance at the Philippines on the Asian front. It was a quite fruitful time for me syncing my view of the world with your more current local insights about the Philippines. Something tells me I might visit while there is still a bit of the Philippines I know left, meaning before 2028, but not this year for sure. Even as I always hope for the best.

      • kasambahay's avatar kasambahay says:

        Irineo, if you do happen to come to philippines for a visit, make sure you are a good swimmer, just in case. you may also need to hire a jetski, just in case too. not to compete with the ill famed jetskier he who would accost the chinese at the west phil sea but never did, but it may be better to drive a jetski than a banca if the weather plays up again and there is flood.

        • I prefer to come during dry season and swim in the sea. I used to be a good swimmer, the muscle memory is still there. Will have to practice again, for sure.

          Scary that even Ateneo campus became a swimming pool recently. Our family went to swim at the real swimming pool there every Sunday before as my mother was faculty. My father never allowed me to go alone there though, as he feared that I might get grabbed by gays, typical UP prejudice about Ateneo.

    • kasambahay's avatar kasambahay says:

      godspeed, joey nguyen. stay safe and keep warm.

  8. Karl Garcia's avatar Karl Garcia says:

    I asked Chatgpt some pertinent prompts

    It looks like you’re asking for information about TESDA, DepEd, CHED, DOLE, DICT, and their AI policies. Let me break it down clearly for you:—1. TESDA (Technical Education and Skills Development Authority)Mandate: Vocational and technical skills training in the Philippines.AI Policy/Initiatives: TESDA has been exploring AI in:Online skills training and e-learning platforms (TESDA Online Program).Using AI for adaptive learning systems to personalize skills development.No formal AI ethics policy yet, but it’s part of digital transformation initiatives.—2. DepEd (Department of Education)Mandate: K-12 basic education.AI Policy/Initiatives:Promotes digital literacy and AI awareness in the curriculum.Has issued guidelines for safe and ethical use of AI in classrooms, especially for students and teachers.DepEd’s AI policy is largely advisory, focusing on:Responsible AI use in teaching.Encouraging teachers to use AI tools for lesson planning and grading.—3. CHED (Commission on Higher Education)Mandate: Tertiary education governance.AI Policy/Initiatives:CHED has encouraged universities to integrate AI courses in STEM and digital literacy programs.Released guidelines on AI in research and academic integrity, addressing:Use of AI in student papers and theses.Ethical AI applications in research.Emphasis on AI literacy for both students and faculty.—4. DOLE (Department of Labor and Employment)Mandate: Employment, labor policies, and workforce welfare.AI Policy/Initiatives:Focused on AI in employment forecasting and skills mapping.Uses AI in job matching platforms (PhilJobNet).Currently, DOLE emphasizes reskilling and upskilling programs to prepare workers for AI-driven industries.—5. DICT (Department of Information and Communications Technology)Mandate: ICT development and digital transformation.AI Policy/Initiatives:Leads Philippine AI strategy and policy frameworks.Developed the Philippine AI Roadmap covering:AI ethics and governance.AI in government services.AI research and talent development.Promotes responsible AI deployment in public and private sectors.—Summary TableAgency AI Focus / PolicyTESDA Skills training, adaptive learning, e-learningDepEd AI literacy, ethical classroom use, teacher supportCHED AI in higher education, research integrity, curriculum integrationDOLE Job matching, workforce AI readiness, reskilling/upskillingDICT National AI strategy, AI governance, public sector AI deployment—If you want, I can also compile the latest official AI policy documents or circulars from each agency so you have direct references.Do you want me to do that?

    • Karl Garcia's avatar Karl Garcia says:

      and a follow-up

      Certainly! Here’s a synthesized overview of the current AI policies and guidelines relevant to government agencies in the Philippines, focusing on the roles of the Department of Information and Communications Technology (DICT), the National Telecommunications Commission (NTC), and other key entities:—1. Department of Information and Communications Technology (DICT)Role: The DICT is the primary agency responsible for the development, implementation, and regulation of information and communications technology (ICT) in the Philippines. It oversees the National Telecommunications Commission (NTC) and the National Privacy Commission (NPC). AI Guidelines:In April 2024, the DICT, in collaboration with the Civil Service Commission (CSC), released a draft Joint Memorandum Circular (JMC) outlining the principles and guidelines for the ethical and trustworthy use of AI in government.The guidelines emphasize transparency, accountability, fairness, and human rights protection in AI applications within government agencies.Stakeholder consultations were conducted to gather inputs and ensure the responsible use of AI technologies in public service. —2. National Telecommunications Commission (NTC)Role: The NTC is an attached agency of the DICT responsible for the regulation and supervision of all telecommunications services and radio and television networks in the Philippines. AI Guidelines:As of now, the NTC has not issued specific guidelines or policies directly addressing the use of AI within its operations or the telecommunications sector.However, given its mandate and the increasing relevance of AI in telecommunications, it is anticipated that the NTC will align with the DICT’s forthcoming AI policies to ensure cohesive governance across the ICT sector. —3. National Privacy Commission (NPC)Role: The NPC is the agency tasked with ensuring the protection of personal data in the Philippines, enforcing the Data Privacy Act of 2012. AI Guidelines:In December 2024, the NPC issued Advisory No. 2024-04, providing guidelines on the application of the Data Privacy Act to AI systems processing personal data.The advisory emphasizes principles such as transparency, accountability, fairness, accuracy, and data minimization in AI applications.It requires personal information controllers (PICs) to assess their operations for personal data processing, adopt governance mechanisms, and ensure data subject rights management. —4. Civil Service Commission (CSC)Role: The CSC is the central personnel agency of the Philippine government, responsible for the establishment and administration of the civil service. AI Guidelines:The CSC has been actively involved in consultations regarding the ethical use of AI in government.Chairperson Karlo Nograles highlighted the importance of raising awareness about the benefits and risks of AI, promoting its proper and ethical use within government offices, and outlining penalties for non-compliance with both overarching guidelines and internal policies. —5. National Economic and Development Authority (NEDA)Role: NEDA is the country’s independent economic and planning agency, responsible for formulating development policies and plans. AI Guidelines:In February 2025, NEDA released a Policy Note on Artificial Intelligence, which serves as a primer on AI integration within the Philippine development strategy.The policy note discusses the current AI landscape and provides directions to maximize AI’s transformative potential, including recommendations for expanding digital infrastructure, strengthening data governance, and developing human capital. —6. Commission on Elections (COMELEC)Role: COMELEC is the constitutional body tasked with administering elections in the Philippines. AI Guidelines:In 2024, COMELEC approved Resolution No. 11064, which regulates the use of AI and social media in digital election campaigns to ensure fairness and curb disinformation.The resolution requires candidates and political parties to register all digital campaign platforms with COMELEC’s Education and Information Department and mandates the labeling of AI-generated content. —ConclusionThe Philippine government is actively developing and implementing policies to govern the ethical and responsible use of AI across various sectors. While the DICT leads the initiative with comprehensive guidelines, other agencies like the NPC, CSC, NEDA, and COMELEC are also contributing to the establishment of a cohesive AI governance framework. As AI continues to evolve, these policies will play a crucial role in ensuring that its integration into government operations aligns with principles of transparency, accountability, fairness, and respect for human rights. If you require further information or specific details on any of these policies, feel free to ask!

Leave a reply to kasambahay Cancel reply