Reaction to AI Developments in the Philippines

This was my reaction speech during the AmCham dialogue on AI governance last Friday, 30 August 2024.

Good afternoon, everyone.

Enough power point, I’m just going old school today and I will read a prepared reaction speech. I’d like to share some thoughts on the recent developments in AI within the Philippines that I have observed both as a staunch advocate for AI adoption – I run a social impact AI company CirroLytix and also an AI watchdog group called Data Ethics PH, and more recently as head of AI and research in IBPAP.

Are we leading or lagging in AI? Depends which surveys you check.

Reports from the Descartes Institute and Oxford Insights in 2023 cast a shadow over our AI landscape. They highlight critical issues such as the lack of infrastructure, a shortfall in skilled human capital, and insufficient policies that hinder AI’s growth in the Philippines.

But that’s only one side of the story.

As Peter (Maquera) shared, LinkedIn suggests that 86% of our knowledge workers are already integrating AI into their daily routines. Writerbuddy lists us as fourth in terms of traffic to AI websites. And Electronic Hub ranks us as the number one country globally for AI-related search interest.

We have also just heard about the wonderful work from DOST as presented by Dr. (Eric) Paringit of PCIEERD, which are evidence, that Philippines is capable of dishing AI solutions. We just need to promote them better.

It’s certainly a mixed bag.

This tells us: while we may face significant challenges in building the foundations for AI, the enthusiasm and grassroots adoption of AI by our people are undeniable.

I think it’s the Filipino spirit—very resourceful and quick to adapt, even without investment!

But this should give us hope that, with the right investments and policies, we can overcome these hurdles and truly capitalize on AI’s potential. If you build it, they will come. In the case of AI, they’re already here, all just waiting for you to build it!

On the legislative front, I want to update what Usec. Fita (Aldaba) shared: there are currently at least eight bills pending in Congress that aim to address various aspects of AI.

Representative Barbers is pushing for the creation of an AI Development Agency, which would serve as a dedicated body to steer AI initiatives nationwide. Representative Tan, on the other hand, advocates for a Private Sector Council on AI and a National Center for AI Research to foster innovation and excellence – the latter of which has already been launched by the DTI c/o Usec. Aldaba’s efforts.

Representative Emano wants to create a GenAI in Education Council to maximize GenAI in schools. Representatives Atayde and Suan, focus on labor protection—acknowledging the potential disruption AI might cause in the job market. Representatives Revillas is looking to empower the DICT to take a more active role in AI governance, while Representative Villafuerte is addressing the ethical risks of deepfakes by proposing penalties for deepfake non-disclosure.

These legislative efforts highlight the complexity of AI governance. It’s not just about fostering innovation; it’s about ensuring that this innovation benefits everyone while minimizing risks and protecting the most vulnerable in our society.

There was a recent Bloomberg article talking about the threat of AI to BPOs. I want to highlight a common misconception: AI may indeed automate tasks, but it transforms jobs. This is like using Waze to get around vs. using a self-driving car. Same data, two different applications!

Now we have yet to hear from our IT-BPM sectors about job displacement – but we’re hearing some noise coming from creatives and IT – demand for graphics, animation, copywriting, and low level dev work seem to be slowing down.

On the other hand, we have many stories of augmentation. AI led to a 14% improvement in productivity in call centers, 30-40% amongst less skilled agents, and a 43% boost in consultants’ performance. In education, AI is making a significant difference, with less skilled students achieving two to three times better outcomes.

The theme is the same: AI does not necessarily replace humans, but helps less skilled performers catch up with their peers

But we must be cautious. There is a number that IT people hate hearing: McKinsey’s rate of failure for digitization projects: 70% (69%) end in failure—not because of the technology itself, but due to a lack of skills, poor design, and inadequate user experience. This underscores the need for a balanced approach where technology is paired with proper training and capacity-building. One BPO company laid off 20% of their staff last year, because their client wanted to migrate to AI bots. One year later, the project had failed and they are now struggling to rehire the staff they lost.

I’ll now wear my IBPAP hat. In response to these challenges, IBPAP is pushing a framework called 4Es focusing on Education, Engineering, Enforcement, and Ethics. We will be sharing more details on this framework soon, but this approach will ensure that our IT-BPM’s approach to AI development is not just about advancing technology but also about building a supportive ecosystem. IBPAP’s talent framework also echoes Peter’s point, we need an all-of-labor approach to AI workforce: Builders, Users, Planners, and Trainers of AI.

Agnes (Gervacio) mentioned inclusion earlier. IBPAP’s emphasis on technology inclusivity will be particularly noteworthy. We define AI inclusivity as: AI Access, AI Literacy, and Trust in AI. AI should be accessible to all, everyone should be AI literate, and we need policies that allow AI to build trust amongst citizens to ensure that the benefits of AI are widely shared.

To conclude, the developments in AI in the Philippines present us with challenges but even more opportunities. The contrasting reports on our AI readiness remind us that while there is much work to be done, the enthusiasm and adaptability of our people give us a strong foundation to build on.

Our legislative efforts, though still in progress, show that we are beginning to take the necessary steps to govern AI wisely. IBPAP’s 4E Framework, will help the IT-BPM sector lay the groundwork for an inclusive and ethical AI future.

Earlier Usec. Fita indicated, almost apologetically, that the DTI had a separate roadmap for AI as opposed to Dr. Paringit’s DOST AI roadmap. I want to say this is something we do not need to apologize for. We need more roadmaps rather than less. Remember even the UN has 17 different approaches to sustainability. Remember that the more tickets you buy, the more chances of winning. But you have to buy the tickets.

Final words: We also tend to judge technology by the present, and less on the future. This tech is changing fast. Let’s stay informed, remain adaptable, and work together to ensure that AI becomes a force for good in our country.

Btw, in case you were wondering, I had my computer open the whole time today. ChatGPT helped me prepare this reaction speech by listening to the speakers earlier.

So I’m not that old school after all.

Thank you.