Hindi natin namamalayang araw-araw pala nating ginagamit ang artificial intelligence. Malaking tulong ba ito sa atin?
Category
🗞
NewsTranscript
00:00 [Music]
00:10 Good day, I am Rosanne de la Cruz and this is Reforma.
00:14 In our program, we will focus on the institutions and agencies in the three branches of the government.
00:21 Are they running well or do they need to be changed to make their services better to the Filipinos?
00:28 In our episode this week, we will discuss Artificial Intelligence or AI.
00:35 Recently, the University of the Philippines released guidelines on the responsible use of Artificial Intelligence.
00:44 According to the guidelines, while AI has its advantages including the potential to transform education
00:50 by enhancing personalized learning, increasing student engagement, and improving education management,
00:56 the rapid development and adoption of AI has led to increasing risks and concerns.
01:02 The United Nations itself talked about this and the possible consequences of using it.
01:08 UN Secretary General Antonio Gutierrez said,
01:12 "AI will have an impact on every area of our lives"
01:16 and advocated for the creation of a new United Nations entity to support collective efforts to govern this extraordinary technology.
01:25 In the Philippines, various government agencies are preparing to help the Filipino workforce.
01:33 But what is Artificial Intelligence or AI?
01:37 Will it help or will it be harmful?
01:40 In our program, we will help our viewers understand how and where AI is used in our daily lives.
01:49 We will have with us Ms. Casey Adviento, an expert in cybersecurity, specifically in risk management and compliance.
01:59 She is currently receiving a master's degree in AI at Queen's University in Canada.
02:08 Good morning, Ms. Casey.
02:11 Good morning to the Philippines. Good morning to you too.
02:18 Ma'am, let's start our discussion on AI.
02:22 Let's explain in layman's terms, what is Artificial Intelligence or AI?
02:28 Artificial Intelligence is like a branch of computer science.
02:35 I know we all have computer science in different universities in the Philippines.
02:42 But the difference is that it replicates our human intelligence.
02:49 We are creating algorithms to teach the machine to mimic human thinking.
02:57 So, it mimics our thinking as humans.
03:05 We allow these machines or computers to perform tasks or do things that were previously required by human intelligence.
03:20 It can be visual perception, speech recognition, or decision-making of a person or translation.
03:33 That's AI in a nutshell.
03:36 Ma'am, just a little bit of history. When did the concept of AI start?
03:42 That's predated my time on earth.
03:46 It's been a long time.
03:49 If we look at the 20th century, we can say that it is Alan Turing and the World War.
04:00 Also, CAPTCHA is still being used today.
04:05 CAPTCHA means the images that we use for authentication.
04:11 When you log into your accounts, it means it's a completely automated public Turing test to tell computers and humans apart.
04:21 That's the meaning of CAPTCHA.
04:23 It's been a long time, but the concept of AI itself could be associated to Charles Babbage,
04:34 which is the father of modern computing.
04:36 But let's not forget that the woman who helped Charles Babbage with this concept, Ada Lovelace,
04:45 was the one who said that this could be something bigger than what we could expect.
04:53 Ma'am, there are a lot of studies and information that have been published about AI.
05:02 For you, what is the most important information that a person should know about this, as in number one?
05:10 I think the most important thing we should know about AI is that it will be a big part of our daily lives.
05:20 Maybe in North America, in this part of the world, it's more on our daily lives, or being affected on it.
05:29 Probably, it's a matter of time.
05:33 We also need to know that artificial intelligence is based on data that we have in the World Wide Web or language models.
05:50 So, AI is made by humans. It can have bias or what we call here unethical practices of AI that we should study.
06:10 That's also the reason why I'm studying AI for my Master's.
06:21 Because this is a very important future of the human race, I guess.
06:28 In our daily lives, Ma'am, where do we use AI?
06:37 AI has been around for a long time.
06:42 I would say that if we watch Netflix, AI is like an algorithm, right?
06:49 As I said earlier, you can see if you go shopping, and you can see if there are recommendations.
07:00 That itself is an algorithm powered by machine learning.
07:05 Machine learning is a tool within the AI. It's a subset of what it could do.
07:12 What you can see on Netflix, the recommendation of percentage, is an example of how AI uses it in our daily lives.
07:25 You can also see in cars that have enabled recognition or auto driving, autopilot is also AI enabled.
07:40 Ma'am, what is the advantage of AI in this time? How much do we need it?
07:49 There are so many advantages, I guess, in artificial intelligence.
07:55 If we look at our daily lives, there are services that are 24/7.
08:02 Let's say, around the clock, there needs to be a service.
08:08 It can be a call center, a factory, or a branch, or security.
08:19 Customer support is really one of the top that we could use artificial intelligence and chatbots.
08:28 We don't need to stretch the hours of people working in the Philippines.
08:40 I know there are BPO industries that have 24/7 operations and graveyard shifts.
08:52 Ma'am, are there things that AI can do that we don't even realize that it's AI that does it?
08:59 The recommendations I mentioned earlier when shopping, I know that on Amazon,
09:09 you bought this, maybe you would like this.
09:12 Behind that is artificial intelligence. It's what you call lookalike model.
09:22 In math terms, we translate it, not math, but statistics.
09:28 It's called cosine similarity.
09:31 If you put what you bought in a cart, you paid for it,
09:39 it will take your data, it will study what you liked, what you viewed,
09:46 and then it will recommend things to you, what you like,
09:52 or what you would like from that model.
09:57 So, e-commerce is one of them.
10:00 Actually, it's really there in applications, in phones.
10:06 Sometimes, you just look at it, and when you view it, the suggestions are there.
10:13 They were telling before that it's like someone is following me,
10:21 someone is looking at me, my phone knows what I'm looking for.
10:30 I think that one is a combination because that's another technology that we will be talking about,
10:37 but also could be powered by AI.
10:39 If you feel that it's surveilling what you're doing,
10:45 AI could make it more advanced or more faster to understand what you actually like
10:52 because tracking is another technology.
10:55 Tracking of your pixels on your phone, what you viewed, what you watched.
11:00 But here, there are things we call human error or human misdeeds.
11:10 Will artificial intelligence reduce human error?
11:14 One of the aims of why now is companies within North America
11:21 are pushing for artificial intelligence as revolution,
11:25 it's also because of the minimization of human errors
11:33 because the AI system is designed to reduce human errors.
11:40 It can be consistent because they are machines, they have the consistency to do things.
11:47 There are no longer human operators in machinery like what we are currently doing right now, I guess,
11:59 in some industry.
12:01 So, those who are leading in this revolution,
12:05 I think that's their benefit that they don't make too many mistakes.
12:15 Ma'am, I believe that all sectors are using or will eventually use AI.
12:23 But what is the most used tool?
12:27 What is the industry that has the most used AI?
12:32 I would say there are many.
12:36 I don't know what I would put first.
12:39 Maybe, the top three or the top five are healthcare,
12:49 second, cyber security, third, education.
12:54 For me, this is a personal perception or point of view, my own point of view.
13:04 Those are the top three that could be really life-changing and could be a help
13:12 to push humanity further.
13:15 What specifically for healthcare?
13:19 So, this is my favorite topic.
13:25 AI that aids patient diagnosis is one of them.
13:32 So, if you have...
13:35 Because now, how we diagnose sickness, it could take days, weeks, months, years.
13:46 Just take for example, like cancer.
13:49 Or I think there is also a technology that pushes that through to make it faster.
13:53 But AI could make it more faster to diagnose you.
13:56 It would help you because of the large or the capacity to do things, to analyze things faster.
14:03 Second, treatment suggestions based on patient data.
14:12 Third, managing...
14:17 This is one of my favorites under healthcare.
14:22 If you have seen managing pain, it could be one of them.
14:29 Based on their patient likes and dislikes.
14:32 Because that's another one.
14:33 There are AR/VR, which is a different technology, but based on the person's data,
14:40 it will be placed in AR/VR on how would the person feel if it's in this environment.
14:46 Then they will see it in AR/VR.
14:48 So, it's all connected.
14:49 So, that's one of those.
14:52 You know, those are good news.
14:53 But I hope it will be accessible to all public.
14:58 If ever AI innovations in healthcare came here in the Philippines,
15:05 because that's what we need.
15:07 Ma'am, when it comes to cybersecurity, how is it helping?
15:12 I don't know if scamming in the Philippines is rampant when it comes to texts or calls.
15:22 What it can do in cybersecurity, there are many.
15:26 Mostly network monitoring, fraud detection, making sure that, you know,
15:34 the systems are diversified, like there are backups.
15:39 But the thing is, in cybersecurity, it's not just the technology,
15:45 but also the human emotion mimicking it.
15:48 So, it can be pros and cons because it could make things faster to detect scam, to detect fraud.
15:58 The other side of the story is the malicious people,
16:02 or the people who are looking for scammers or fraudulent people,
16:12 they could also use this as a tool to expedite their operations.
16:21 So, it works really hand in hand.
16:23 So, it needs both, you know.
16:28 Ma'am, this AI can make decisions on its own.
16:33 Is there anything we can expect, a right decision that came from this,
16:38 since there is no emotion involved here, unlike humans making the decision.
16:46 Yes, it can make decisions based on its data.
16:55 Because it will not operate, it will not, I think it can create based on what you feed it.
17:06 That's all.
17:07 So, it's hard to say out of the blue because what you're saying is sentient,
17:13 it has its own thinking, which is the companies, the tech companies in the world,
17:22 is trying to create that advanced AI even further.
17:27 So, which is the other billionaires in the world are trying to stop that to happen
17:33 because they know we're not ready yet.
17:37 And yes, Ma'am, go.
17:42 It can make decisions on its own once we have produced something like that, that's close to sentient.
17:52 And yes, because this artificial intelligence is programmed,
17:57 it cannot come up with out-of-the-box solutions to new problems.
18:05 Yes and no.
18:07 Yes and no.
18:09 Yes, because it's based on what we've studied or what we've fed,
18:15 what we've put data to analyze the algorithm of AI.
18:21 Although, it can learn.
18:24 That's what we call generative models or GANS,
18:30 some AI models which are currently building here in this side of the world
18:35 or probably there are also some in other parts of the world,
18:39 but I'm not very aware of other parts.
18:42 It has real-time information that it takes in.
18:47 Basically, it's generative from the word itself.
18:52 So, it learns from what you teach it.
18:57 So, it can have an interaction between you.
19:01 For example, in the classroom model, let's say I'm the teacher and you're the student.
19:11 The concept is you're the AI that I'm teaching and I'll tell you how to do this
19:22 and then you will learn from me.
19:24 Based on the data that you know, you will adjust based on what you learned from me.
19:31 If it's other people that will teach you,
19:36 you can form your different perception or different point of view
19:44 based on the people who you learned it from.
19:49 Oh, I see.
19:51 When it comes to one of the things that AI does,
19:59 for example, in the routine jobs, like in the factory,
20:07 how does AI help in this kind of industry?
20:18 I'm smiling, sorry.
20:23 Because in the routine jobs, like customer support, manual labor jobs,
20:29 I would think it would help them, especially in the dangerous parts.
20:33 For example, in the factory where the machinery is a bit dangerous,
20:38 like human risk, high-level human risk,
20:41 what they could do is, let's say, AI robots that can be developed could help
20:48 replace those, not just replace, but help, aid the human to do the job.
21:00 Because there should be human intervention as of now in the routine jobs.
21:12 When it comes to the business sector, there is what is called applied AI.
21:17 How does it work?
21:22 It depends on the industry.
21:26 It can be a specific business problem, business platform.
21:32 As an example, marketing.
21:36 It can generate marketing content for you,
21:39 but again, with the help of human intervention.
21:42 Because human touch is still needed within marketing.
21:47 But it can help on generating that content.
21:52 And then, there's planning on how to execute marketing.
21:56 In other business units, finance is one of them.
22:02 And then, customer service bots.
22:06 What we see right now in chats, they have a bunch of FAQs, companies.
22:15 And then, it's like you have a customer who will just go to your website
22:21 and type what they need as a service company,
22:26 and the chatbot will just respond to you.
22:30 So, it will be helpful for the business to shrink down the time
22:37 and also the resources within the company.
22:41 So, it chunks.
22:45 What AI can do in a business depends on what kind of business it wants to happen
22:55 and where it can be implemented.
22:57 Because different sectors of the business or different facets of the organization can use it.
23:04 Earlier, we talked about the advantages and the help that AI can do.
23:10 Let's talk about the disadvantages.
23:13 What can we face when using AI?
23:18 Okay.
23:20 I know that the technology, artificial intelligence, is really promising.
23:27 But it requires a lot or a significant amount of money or investment,
23:33 hours as well included, but the talent,
23:36 which is the people who will build it.
23:41 Also, the infrastructure is a disadvantage because those are expensive in terms of dollar amount or peso amount.
23:50 Another one is, since the tech sectors are robust within the North American region,
24:01 we really felt it. The AI revolution here was a shockwave when it first launched.
24:08 There was fear that people will replace existing jobs.
24:17 An honest answer is, yes, it can replace jobs.
24:23 But it could aid you.
24:28 Like what I said earlier in the factory,
24:32 the examples of dangerous places where AI can help,
24:38 I would still think that human intervention or human monitoring is also needed in these times.
24:48 We are not yet super advanced AI.
24:52 We're still at the very beginning.
24:55 So, in this generation, we're still in the process of building AI,
25:02 which requires a lot of money, a lot of time, a lot of infrastructure.
25:08 So, those are the disadvantages that I see now.
25:12 Ma'am, here in the Philippines, the preparation of our government agencies is continuous.
25:19 The DOLE, the Department of Labor and Employment, is there.
25:25 They are preparing.
25:27 It's like there are preparations for the AI revolution.
25:32 The DEPED, of course, in our education,
25:36 like what you said earlier, that AI is one of the tools used.
25:42 When it comes to DILG, the private sector,
25:45 they are really training underskilled women.
25:50 What can you say about this?
25:52 I think that's very much needed.
25:56 So, I think what I mentioned earlier, the ethical AI and the biases,
26:01 the challenges are different.
26:02 So, I didn't put that in the disadvantages as well
26:06 because our training data is biased.
26:08 The world's data is biased.
26:10 So, to start off, we have to know what artificial intelligence is.
26:17 I think what the government is doing is a really good step on making sure we know how it will impact the industry,
26:29 impact society.
26:31 Because it's way too…
26:34 The change, I guess, is drastic.
26:37 Because, like I said earlier, it's a shockwave that the CHAT-GPT was released,
26:47 which is text-based.
26:49 So, it's been a while since AI was introduced,
26:53 but then what the Philippine government is doing is actually thinking ahead.
26:58 I think they are ensuring that there are agencies in the government that are working together with different industries in different parts of the world.
27:13 Because we cannot be left behind.
27:17 Because if we are left behind, I think we will catch up more.
27:27 And we might be affected by the economy.
27:33 Yeah, I think that's helpful because we have digitization efforts here in the Philippines.
27:42 All of our agencies in the Philippines are catching up to adapt to our digitization.
27:51 But, there are dangers to artificial intelligence.
27:58 Because recently, in education, there are students who used the CHAT-GPT
28:07 when it comes to hand-writing manuscripts,
28:10 or scholarly documents, essays,
28:19 which are passed on to teachers using AI.
28:23 Which I think is being regulated.
28:25 That's why UP said that this should be regulated.
28:38 What I can say is, I think,
28:42 to the students or professors of universities,
28:50 they know their students,
28:55 I'm saying this in TMTCS,
28:58 when I was talking in Manila Times College of Subic about AI.
29:05 The professors and other staff members of universities know the ability of students.
29:15 I know that they use it to cheat.
29:20 And the detection for cheating in universities,
29:26 I know it's hard to find a cheat detector now.
29:32 But bottom line, the professors know their students.
29:36 But I would think we should go further than that.
29:40 We should go further than just cheating itself.
29:43 Because in Europe, there's a draft on AI law.
29:50 So, the draft of AI law, we should think about, really,
29:55 the people, the society,
29:58 you said earlier that you feel trapped.
30:04 So, AI law prevents, not just prevents,
30:09 but sets limits to what AI can learn from the people.
30:14 So, when they get the data from different set of people,
30:24 different set of subpopulations,
30:28 it has a different impact.
30:31 It can change your way of thinking.
30:37 If we heard about the Cambridge Analytica issue,
30:42 that's AI powered by it.
30:46 So, what I can say is,
30:50 if we're thinking about university and students on cheating exams,
30:56 take a step further, even further,
31:01 to protect the citizens, the population,
31:07 and all their data is, you know,
31:11 protect them from businesses who will take advantage of them.
31:16 Because that's life.
31:21 The tool itself could change the way how you look at someone,
31:27 at something, makes you buy things,
31:30 makes you watch something,
31:32 that you don't have the persuadables.
31:37 So, it's the danger.
31:40 That's included in the drafts of law in different parts of the world.
31:48 AI law in EU is being drafted.
31:54 In the US, New York is trying to do something as well.
31:59 So, I think, take a step further,
32:03 like stretch it more.
32:07 Ma'am, we've talked about the dangers of AI.
32:12 What are the privacy and security issues,
32:18 concerns that we can discuss here in AI,
32:22 that we can experience?
32:25 So, there are tools, I know a lot of people are using this now,
32:34 Chad G. Beatty is one of them.
32:36 But there are so many AI tools.
32:38 Or the ones you see on the internet or your apps on your phone,
32:44 they will say it's AI enabled.
32:46 In the Philippines, those who want to generate images of themselves
32:52 in a different ethnicity, they do that.
32:55 But you have to be aware that the data you are sharing,
33:00 or the pictures, data considered as photos,
33:04 or personal information that you share to apps, companies,
33:10 you share it freely.
33:12 They can use it for free and you just gave it to them.
33:16 So, by itself, you're giving away your privacy.
33:23 So, that's a really concern because they are using their models
33:29 to learn people, to learn the population.
33:34 So, I think, not all of them are caught yet,
33:40 so that they won't give data freely to different apps or companies.
33:45 But we have to think that we're just giving away our privacy.
33:50 Security concern, I've already said that.
33:54 You can see that when you post a lot of things on social media,
34:02 AI can, I guess, it can help malicious people.
34:11 I've said it earlier, they can help malicious people
34:13 to expedite their crimes.
34:17 So, imagine if you're sharing your information freely
34:22 and they can just generate, for example,
34:26 just an example, your social media.
34:30 You posted your birthdays, let's say, you have a photo of friends
34:39 and then your ID is there somewhere,
34:42 the photo is being captured.
34:44 That itself is a privacy issue.
34:46 AI can regenerate the photo even if it's blurry.
34:49 So, those things, right?
34:52 If you haven't thought of it, that's how it works.
34:55 So, we have to be aware of that.
34:57 With all the things that AI can do,
35:00 with all the things that AI can help us,
35:03 what can Filipinos do so that they won't be,
35:08 well, not just the Filipinos, but what can people do
35:12 so that they won't be obsolete when it comes to AI?
35:18 That's a good question.
35:20 I really like that question because
35:22 all the conferences, the talks that I attend,
35:32 those are the questions that really reflect
35:37 that, "Okay, let's not be afraid."
35:39 Maybe it's adoption because people are a bit afraid
35:44 that their job will be replaced.
35:46 Think about artificial intelligence to aid you.
35:51 They will help you with your daily activities.
35:55 So, they can also help you study the things you want to study.
36:02 For example, if you want to study Japanese language,
36:06 it can help you do that.
36:08 Or English language, like proper writing on a newspaper
36:15 or things like that.
36:16 You can order AI to make you an entire curriculum
36:26 so that you can study that.
36:28 So, think about this as a tool or as an assistant
36:31 that could help you with your daily activities.
36:34 Actually, ma'am, they say that you shouldn't be afraid of AI.
36:39 You should be afraid of people who can use AI.
36:45 So, that will be your enemy, not AI itself,
36:49 but the person who can use AI,
36:53 who can and can use AI to their advantage.
36:58 Yes, that's true.
37:00 Because I think when we say that people use AI,
37:06 they immediately think of, because this is a branch of computer science,
37:11 they would think they need to code.
37:13 They would think that they need to study first.
37:17 No, you don't need to study first because there are text-based,
37:23 like BARD by Google, there's Bing,
37:26 and the things we do,
37:30 maybe everyday chats with friends,
37:36 chats in forums, or other things like that,
37:42 you can talk to an AI that's text-based that easily.
37:48 And also, you can generate images through text,
37:53 which is your mid-journey or dolly.
37:56 So, you don't need to learn how to code,
38:00 you don't need a computer science degree to use AI.
38:09 You don't need a degree at all to use it,
38:14 just like how people use Facebook or Instagram or TikTok.
38:18 It's the same idea.
38:21 Thank you so much, Ma'am Casey, for enlightening us.
38:26 When it comes to AI, Ma'am, do you have a last message for our fellow countrymen,
38:31 for our viewers, about what they need to do with AI,
38:38 what is your message to them?
38:42 Yeah, so AI can be good and bad, and this would depend on the people.
38:47 But let's first see what AI can do.
38:53 Maybe we're afraid, because it's scary when we don't know,
39:00 or we don't have any knowledge about AI.
39:03 But as I was saying, it's just as if you are talking to a friend or a chatbot,
39:12 similar approach.
39:13 So, try it on and use it as an assistant for your day-to-day life
39:21 or anything that you could think of, even a recipe.
39:27 And one of my messages is,
39:32 let's not forget that artificial intelligence has biases.
39:42 There are unethical practices on how to use it, and we have to acknowledge that.
39:48 And when we know that we are subject to bias, we can protect ourselves.
39:57 There you go.
40:00 Thank you so much, Ma'am Casey, for accepting our invitation.
40:06 AI, from its core, is an artificial intelligence.
40:12 It is expected to help or add capabilities to human intelligence,
40:16 or the natural intelligence that comes from humans.
40:19 But according to experts, we can't expect everything here,
40:24 because humans can't replace it.
40:26 During digitization, it's important to write our eyes on this thing.
40:31 But we should always remember that all our actions have a responsibility,
40:37 so we should be more observant and know about things.
40:42 Thank you so much, Ma'am Casey, for accepting our invitation.
40:46 Thank you so much.
40:49 This is where our discussion on hot issues
40:54 that affect the public service of government institutions and agencies.
41:01 In the change for the future, let's push for reform in the government.
41:06 Thank you so much.
41:08 [Music]
41:15 [Music]
41:16 [BLANK_AUDIO]