Google Cloud Speech-to-Text API (Chirp) + Gemini 2.5 Flash Non-Thinking. Treat it as a searchable index of what was broadcast, not a quotation record.
00:00:00Valencia, a day later at the same time.
00:00:03Tickets are already available for purchase through
00:00:07web platforms with payment facilities so that their
00:00:10fans can enjoy a great evening of romantic
00:00:12salsa.
00:00:15Friends, that's all for this news update,
00:00:19I always invite you to connect to our contact points on your screens,
00:00:23our coordinates, our
00:00:25and I say, well, look, what website, and of course our new YouTube channel. I want to invite you
00:00:28from 9 p.m. to the... Stellar broadcast. I'll be waiting for you.
00:00:48right? So of course, I am taking
00:01:00Health goes beyond urgency, it
00:01:04is built with evidence, with questions that others avoid, very
00:01:08important, a respectable myopia years
00:01:11ago we understood that medical information should not
00:01:14scare, it should protect, we compare
00:01:17studies, we consult experts and we filter
00:01:21everything to keep what is really important for your
00:01:23life, beyond replicating alarms we give you
00:01:27certainties hand in hand and the meeting. The laser no, because
00:01:31taking care of your well-being more than a
00:01:33both sellers and service is a commitment that we renew every day,
00:01:49in a planet that never sleeps, you must pay attention to
00:01:52every event, what is moving in the
00:01:57this is not an IT world today, art that inspires, technology
00:02:00that surprises, news that impacts and things
00:02:03that only happen
00:02:06once, from Asia to America, from
00:02:10Europe to Africa, it is a global experience
00:02:13with all its faces and
00:02:16it seems it will be very different from what all its voices, a weekly date with the whole
00:02:19planet, view the world Saturdays
00:02:223 p.m. on
00:02:24Globovisión,
00:02:33Jesús Pregunta, a mixed program, informative and
00:02:35Montiel, professor at opinion with elements of language, health, sex and violence, suitable for
00:02:39all audiences, a nationally produced
00:02:42independent program, certificate number 35,906.
00:02:47to demonize it, saying, look, it's going to
00:02:52think about, the data that I have within
00:03:05Good afternoon, at this time, greetings to those
00:03:09who tune in to Globovisión, my name is Jesús López and this is your program
00:03:11Jesús Pregunta, today we are going to talk about something that
00:03:15you are surely already using and everyone here in the
00:03:19studio, it is called artificial
00:03:22intelligence, for that we have Gabriel
00:03:25Montiel, AI expert, professor at the Catholic
00:03:29University. in this area no yes for the
00:03:32invitation well let's start talking about AI as an
00:03:35instrument that is already in people's homes, that is,
00:03:39from mobile applications to message
00:03:42writing to whatsapp, no yes, AI as
00:03:45I always say in class is a snowball that will not
00:03:49stop growing, it is already very easy to access
00:03:53is, to delineate the veins of it, it is free, it is also on everyone's cell phones and
00:03:56well now it's about how we are going to use it, what we are going
00:04:00to use it for and how it will help us, right? I think that is the
00:04:04powerful thing about the conversation in terms of artificial intelligence,
00:04:08we have seen forums on
00:04:11the internet with people already talking that we even have to put a stop to it, is
00:04:17invoices you have in the that possible? Yes, well, right now there is a debate because the owner
00:04:21president of Antropic said
00:04:24free, okay? Why? Because the that in 6 to 12 months there would have to
00:04:26be governance because in 6 to 12 months there would already be
00:04:30complete access to AI on the internet so how were we going to manage
00:04:33those policies, right? And look,
00:04:37deny the use of that's what happens there and we do it a lot in corporations when I go into companies is
00:04:41that the first thing to look at
00:04:45is how we are going to govern artificial intelligence, right? And
00:04:49there it is divided into two things: one thing is the AI adoption
00:04:52policy, that is, how people will start using it, what I will
00:04:55use it for, what it will do for my
00:04:58excel, what data I will input, how the company
00:05:02will manage that, how I will train the different employees in that, so
00:05:05like the policies I will follow in those terms, but
00:05:09on the other hand, in terms of that governance, what is the
00:05:13data governance, right? Okay, because one thing is chatting with
00:05:17AI, which everyone has access to now, but another is
00:05:20operating with AI, and operating with AI is, for example,
00:05:24in fact yesterday I was with a massive food corporate talking, right?
00:05:28So how do I make it so that sellers who are selling
00:05:31three products to a client and of course they don't know, they visit
00:05:35many clients, they are, let's say, involved in the
00:05:38process, it suggests, for example, from a portfolio of 30
00:05:42products, that they put these four other products that are also
00:05:45stagnant in inventory, right? So AI can operate there and can
00:05:49make that type of suggestion, but for that to happen, the
00:05:52data must be governed, and the data is where the company's
00:05:56information lives in the systems,
00:06:00this explanation to the company now, but as you ask me, this discussion
00:06:04that is happening now that the president of the United States also came out and so on, it is a
00:06:08controversy, well, it is that governance of AI, but now in
00:06:11global terms. And in global terms it is, look,
00:06:15I have the agents, I have access to the internet, how are we going
00:06:18to govern this in terms of state and so on, right? Because before talking about
00:06:22precisely that, how companies and large
00:06:26corporations have been using AI to, well,
00:06:29perform more, sell more and be more efficient,
00:06:33right? Before talking about that, I would like to talk about these topics that are very
00:06:36particular, for example, humanity
00:06:40versus
00:06:43AI, right? How humanity sees
00:06:46itself in danger or not in danger from your point of view, you know
00:06:49what reminds me of it because I said it in class a few
00:06:53days ago, the student asked me what
00:06:57things AI would take away from me at work and so on,
00:07:01and I told him everything, it will take everything away from you, everything you do it will
00:07:05take away from you, but it will not be taken away by a robot that comes to do it, but by a
00:07:08person who is trained in the use of this type of tool and it
00:07:12is important, artificial intelligence
00:07:15is basically a very fast data processor,
00:07:18okay, something that for us you
00:07:22could take days, months, AI does it in seconds, and
00:07:26I always give the example of the tic-tac-toe game where you put x0,
00:07:29right? Well, how does AI work, I, let's say, I am the AI and
00:07:33you are Jesús, we are playing, you put x, and I, as the AI, have
00:07:36to calculate the mathematical probability of putting the
00:07:40zero, but also of putting,
00:07:42what you will put after I put the zero and then do the entire
00:07:46game in less than one second, microseconds, right? So
00:07:50that speed in taking out and that algorithm in taking out that probability is what
00:07:54makes it process data very fast, so it is not magic, it is not
00:07:57that it comes from a context and a data, so well for your
00:08:01question humanity and AI I think that
00:08:04basically humanity is giving that context so that AI as a tool
00:08:07begins to give us things faster, right? If we could talk about that
00:08:11black box, right, the storage of
00:08:14data, from a global
00:08:18point of view, could we really talk about
00:08:22an audit? That is, do we know what
00:08:25information is being provided to AI from that point? Is AI
00:08:28absorbing all types of information? Well, that's a
00:08:32very good question because look, if today
00:08:35you walk down the street here in Venezuela, right, just to put it in
00:08:39terms of here, and you ask, look, which one
00:08:42are you using? Because everyone is already using it in one way or another, right?
00:08:45So they tell you, yes, I am using Gemini, which is Google's,
00:08:48right? Why? Because it's the free one and it's the one with the most access in
00:08:52Venezuela, okay? But you are putting all the data there in the free
00:08:55part and specifically in the free part it says
00:08:59that everything you put there will be used to train Google's LLM model,
00:09:02okay? So basically we are giving
00:09:06all the data and we also do that in everything we do,
00:09:09browsing on the cell phone is traceable, right? Okay, so
00:09:12there we are giving all the data, so there are some things
00:09:16that happen, some conversations that happen a lot in corporations, how do
00:09:20I protect the corporate data, right? But well, in
00:09:23your question is whether AI has access to that data because
00:09:27we are also giving it to it and if you notice all these
00:09:31companies from China, United States, all of them, it's always something
00:09:36model needs to emulate human language, the more we
00:09:40use it, the more context they will have. on how to train the model, right?
00:09:44So that's why there will always be something free to use, let's go to a commercial break and we'll be back.
00:11:18September 15, was declared by the
00:11:21United Nations in 2008
00:11:25as International Democracy Day, in
00:11:28adoption of the Universal Declaration of Democracy
00:11:31by the Inter-Parliamentary Union in
00:11:331997,
00:11:35a topic now more relevant than ever, with the changes that are
00:11:39taking place in the current global geopolitical
00:11:43recomposition. In Venezuela, Democracy Day is
00:11:45celebrated on January 23, in
00:11:49memory of the fall of Pérez Jiménez's dictatorship that day
00:11:53in 1951.
00:11:56This was the analysis on
00:11:57the carpet.
00:12:06This Sunday in Embassies and Ambassadors, Julio César
00:12:10Pineda talks with Beresit Rigal, CEO of Silicon
00:12:14Valley Venezuela and Christopher Leal,
00:12:17Communications Director about the impact of Silicon Valley
00:12:20Venezuela on the global
00:12:22ecosystem.
00:12:26Sunday 7:30 p.m. on Globovisión.
00:12:57We are talking with Gabriel
00:13:00the Catholic University and well, AI expert,
00:13:03we are going to talk about the good things about AI because it seems
00:13:07that it is very demonized, right? Some sectors would like
00:13:12replace human beings, we have already
00:13:16seen that if it is used very well, it complements
00:13:19human efficiency, yes, from the corporate
00:13:23point of view. In what is it advancing
00:13:26right now with respect
00:13:29to data analysis,
00:13:32systems, a very important topic which is
00:13:35sales, how is it being done? Well,
00:13:39look, when we are going to think about artificial intelligence in
00:13:42terms of systems and and the company, the first
00:13:46thing to think about is data, okay? Because how does artificial
00:13:49intelligence work, I explain it as a factory, right? In the
00:13:53factory there is raw material, raw material enters a machine to give you a
00:13:57product, well, AI is, that raw material is context, it is
00:14:00data that enters some algorithms and probabilities to
00:14:04give you an answer, okay? So the first thing in the line
00:14:07is the raw material, that is, the context and the data and it happens to me a lot,
00:14:11that I am going to enter to
00:14:14do AI architectures, what is AI architecture, that
00:14:18the organization to know where I can put an AI model
00:14:22to help you with certain things, to help you do. bank reconciliations, for
00:14:26example, exactly, let's talk about a specific example, for example,
00:14:29finance, to help you cross
00:14:33purchases with bank outflows, you do reconciliations, to help
00:14:36you ensure that payments that came in are crossed with sales orders or
00:14:42sales department, to help you know what things you need to move from
00:14:46inventories and sell, to help you know the depth of what you want
00:14:49to sell each client you have and by area,
00:14:53what sells more in Venezuela?
00:15:03How the seller has that information quickly, right, and that it does not necessarily depend
00:15:07on him being an expert as a seller over time or on street
00:15:10surveys or door-to-door
00:15:13market studies, for example, right? So of course, it's not about
00:15:17artificial intelligence
00:15:21doing the work and so on, it's about it being a very fast data processor that can
00:15:25help me have things so I can make the decision, so this always
00:15:28has to be seen as a co-pilot, right? You have a co-pilot that is now like a great
00:15:32assistant, the assistant everyone wants, the AI that you want everything,
00:15:36well, how do I make that co-pilot work and that
00:15:39co-pilot only works if the data is
00:15:42organized there, that is the big issue when one is going to
00:15:45implement AI. I am going to build the AI architectures and when I come
00:15:49to see there are no systems, the data is not organized or
00:15:53we are all using an excel here and there, right? So I say,
00:15:57well, the first thing to verify is where that
00:16:00information flows, okay? Where does it originate? It originates because you made a
00:16:04sale, then it originates because you made an invoice, then it originates because
00:16:07you received raw material from some suppliers, so from
00:16:11the roots we have to have, let's say, a
00:16:14digitalization, AI only lives
00:16:17if the digital world exists and
00:16:21it is not something that is
00:16:24not only AI, AI is put on top of a data
00:16:28flow, so we have to think about data first. How do you see the development of AI in
00:16:31Venezuela? Look, I am very enthusiastic and
00:16:34I am also seeing it there, if in fact you go to LinkedIn right now
00:16:38and search for AI, many
00:16:40corporations are looking for AI engineers,
00:16:44AI specialists, AI
00:16:47intelligence specialists, artificial intelligence analysts,
00:16:50because there are two ways, one
00:16:54is chatting with AI that gives you a lot of
00:16:57productivity very quickly, that is, and those are the
00:17:00trainings that normally happen when
00:17:04one enters, right? Like I need to relieve the...
00:17:07situation so one trains
00:17:10human resources, for example, and the finance
00:17:14department, etc., to use chat with controlled licenses and
00:17:17so on for the data, but then you start to do it very quickly, the excels come out very
00:17:21quickly, dynamic tables, graphs, dashboards, analytics,
00:17:24you start to automate things with natural language that before was something
00:17:28like, hey, I need the programmer, I need
00:17:31IT, this is something very important,
00:17:35people's issue, okay, it's your issue,
00:17:38Jesús, it's my issue, it's my mom's issue who goes
00:17:42crazy incredibly doing theology
00:17:44she does things that make me say I can't believe this about Gabriel, how does this process
00:17:48data, right? So that part of chatting relieves a lot
00:17:52of the process, another is operating with AI which is a little
00:17:56more complex in terms of systems, right? But it's already something so that the
00:17:59company works much faster with AI, the
00:18:02competition between companies and AI,
00:18:06right? Because in simple terms,
00:18:09I come and ask a question and
00:18:12suddenly someone else asks a
00:18:15similar question, so how is the competition between companies managed with
00:18:19different information, or is it the same
00:18:21information?
00:18:23Let's remember that information is data and that context is what we have to
00:18:29the company is not available to another company, okay? If another company is
00:18:32using AI in general free and so on, it is the
00:18:36information that everyone has, it will not have the important
00:18:40juice of AI, if it is using AI with its
00:18:43context and with its data, it is too powerful what it will begin to
00:18:46achieve, right? One of the things I say in class is everyone will
00:18:50use it, how are you going to differentiate yourself from everyone else, right?
00:18:54It's the same, everyone will ask, make me a summary of this, and
00:18:58the summary is incredible, that is, it passes it and it's beautiful and it's
00:19:02like resumes right now, all of them are beautiful now, that is, the
00:19:05horrible resumes were normalized, now all beautiful, but well, in truth
00:19:09there is something behind it, that is, it supports what the
00:19:13resume says, right, read what you are executing as another
00:19:16tool, right, we have to give it the criteria and support to
00:19:20that, so how does the competition happen?
00:19:24The competition will be won by whoever best uses their context with
00:19:27AI, point one, and whoever loses their fear of
00:19:31the tool, it is not magic, it is another tool that must
00:19:35be learned to use. There are already companies
00:19:38in Venezuela that are, you
00:19:41would be surprised, but many, yes, this, I think there is a lot of work
00:19:45right now, companies are very active, a president
00:19:48told me, Gabriel, I feel like I'm going to get stuck and I can't,
00:19:51that is, we have to enter this, ah, because I remember that
00:19:55conversation was like, look, let's relieve a specific
00:19:59part of logistics, the routes of the trucks, how they move through Venezuela and
00:20:03so on, so it was like, look, wait, what I am
00:20:06seeing is that I need this for the entire corporation, I also feel that
00:20:10Venezuela currently has an important push, I feel that it will
00:20:13be better, there will be much more growth and
00:20:17also entry of companies and the locals who are here say,
00:20:20well, others can come with a lot of momentum already using it, I have
00:20:24to be already trained for the new market, let's go to a
00:20:27commercial break and we'll be
00:20:28back.
00:21:10My loyal and adored viewers, this Sunday
00:21:14we start with a good party, wait, come
00:21:17more, come more.
00:21:23One thing I will tell you, if there is something in the
00:21:26program that you didn't like, then
00:21:29blame me, blame
00:21:32me,
00:21:34I ask God to give me much health to
00:21:38be able to be with you all the
00:21:40time in the world, today I have time, I have all the
00:21:44time in the world, again we will meet this Sunday at
00:21:487 p.m. in family, with the golden
00:21:52years and with whom else, with
00:21:55papaito,
00:21:57the golden years with Álvaro Cuchillo Durán,
00:22:01Sunday 7 p.m. on Globovisión.
00:22:30We continue with your program, Jesús Pregunta, we are talking with Gabriel
00:22:33Montiel, professor,
00:22:36disinformation and AI, right now
00:22:40we are talking about the times of fake news,
00:22:44times of seeing images that seem
00:22:47very real, that are not real, how is that detected?
00:22:50The society that is starting to use it, well,
00:22:53a very, very good question. That's how misinformation occurs
00:22:57in two ways: one by those who do not know how to use
00:23:00the tool and that training and that
00:23:03misinformation specifically causes fear, that's why the fear
00:23:06and that's also why there is a lot of resistance
00:23:09to change, right? Like I resist the change of this new technology
00:23:13and so on. That misinformation leaves you behind
00:23:16in what is already happening, which is that companies
00:23:20are adopting it, right? But on the other hand, there is also
00:23:23the misinformation that is fact-checking, that is, how do I
00:23:27certify that this fact in which I am publishing a news
00:23:31item, for example, or an opinion or an image, is actually real,
00:23:35okay? And that is one of the things that right now is very
00:23:38complex to decipher because
00:23:42AI is becoming more and more accurate, okay?
00:23:45So one never knows if the
00:23:49information is really accurate or not, there are tools that allow you to know if this
00:23:53was created with AI, right? But there are also ways to make
00:23:56what you write better and it is done with AI, right? Here always comes
00:24:00the issue of ethics and it is one of the things that I, it is the first
00:24:04class of the semester, right, like the ethics of AI,
00:24:07so I always comment, well, is a knife ethical?
00:24:10Well, a knife is a tool, the ethics is given by
00:24:14the use that humans give the knife. If you are going to use the knife to kill someone,
00:24:18that loses all type of ethical criterion, but if you are going to
00:24:22use it to cut food and so on,
00:24:26there is no problem, right? So the use of the tool is given by, the ethics behind the
00:24:30use of the tool will be given by the human behind it, right? And I feel that
00:24:33in the end, Jesús, the truth always comes out, the truth is always known
00:24:37and it is known who wrote it, right? No, so I think we have to be very
00:24:41transparent with that, talking about the topic that
00:24:44he is a professor and also a professor in this
00:24:48AI subject, how does a
00:24:50professor compete in a few years with
00:24:54AI so that the student pays more attention to the professor
00:24:57and doesn't get bored, for example?
00:25:01That's great too. Look, in fact we have had to train
00:25:04professors as well because as I
00:25:07told you, misinformation causes fear and a first
00:25:11shock and many of the professors, in fact my
00:25:14sister in the Dominican Republic gave this type of training to professors, in
00:25:18fact, from school.
00:25:21it is inevitable to use AI, all students are
00:25:24using it, if you
00:25:27AI, you are finished, okay? That is, there is
00:25:31no way around that, it's like if when I was young I had refused to use
00:25:34Google, that is, one researched with Google, it was something
00:25:38very normal, so the professor has to, well, first know the technology and
00:25:42now know that they have to change the ways of evaluation, okay?
00:25:45Everyone will have access to use it and it is very easy, now how did you get
00:25:49to that answer, that is, the report, I know you are going to do it with artificial
00:25:53intelligence, and it will be incredible, that is, it will be beautiful,
00:25:56but I have to go into the depth of the content of how you got to that, that is, it is no longer about seeing the
00:26:00report, it is about how you made the report, okay? So it is about
00:26:04evaluating your prompts, what you wrote, how you got
00:26:07to that, right? So that type of training we have given
00:26:11and also a little bit the activities, right, changing the activities and that
00:26:15is a very important challenge for the professor and
00:26:18I see it every day, there are older people who have been doing this for many
00:26:22years and it's like, wow, you just changed the rules of the game
00:26:28there is is a little bit to see how it works and change the activity a little, right?
00:26:31But don't turn away or
00:26:35stop it because it will be inevitable and no, rather what it gives you is
00:26:39a headache if you don't adapt to it, the
00:26:42algorithm, how
00:26:46to know if the algorithm will at some point, which already
00:26:50happens, but make decisions thanks
00:26:53to the algorithm that is on your phone?
00:26:57Yes, today you make them and in
00:27:00fact in meetings they tell me that
00:27:03I don't make decisions because I don't use social networks
00:27:07and I say, well, it's totally false is, uh, all
00:27:11the world on average in Venezuela we are at 1.5 hours of social
00:27:15network use and we are on the rise and that
00:27:18is all, all ages, that is to say that
00:27:2160 plus is also at 1.5 and above, all decision
00:27:24makers, everyone is using it and no matter how much you don't want it, that
00:27:28influences your decision to see things, etcetera, it influences what you are going
00:27:31to buy, what you are going to think and so on, right? So the
00:27:35algorithm, how will it influence, yes it influences, it will not stop influencing
00:27:39and besides, the algorithm is like the Coca-Cola formula that nobody will
00:27:42know because each company manages its algorithms and it
00:27:45is like, uh, it is the real asset they have and it is something that
00:27:49you will not know, right? But if you are using the
00:27:51tool, uh, without a doubt you will be,
00:27:55uh, that will influence you. You
00:27:58who are normally in the case of...
00:28:02You as a professor of this, of this
00:28:04area, uh, how do
00:28:08you predict the next few years, that is, we could talk about
00:28:112030, there is talk of 2030 as the year where
00:28:14technologically
00:28:18we are experiencing now.
00:28:21Yes, totally, and I will give you two concepts to
00:28:25talk about narrow AI and general AI. What we live today is
00:28:28narrow AI, that is to say, it is
00:28:32tools made specifically to do something very well,
00:28:36the Tesla vehicle is made not to crash it, the AI is to not crash and
00:28:40drive the vehicle well, right? Google Maps that marks red when
00:28:43there is a queue is specifically for that, to mark the data on
00:28:47traffic, ChatGPT to give you the answer, yes
00:28:51there are many tools right now to make video, to make
00:28:54images, to even make songs, okay, that is
00:28:57narrow AI, narrow AI will lead us to general AI which is that
00:29:01idea of the humanoid robot self-deciding and
00:29:04self-creating, right? Uh, we want to get there, but for that
00:29:08we have to emulate human language correctly, but that will come out, well,
00:29:12every day a new tool comes out and infinite ones will come out, why?
00:29:16Because we must learn how we do things and that must
00:29:19improve, when will we get there? We are not in general
00:29:23AI today, close, well, I could tell you that I
00:29:27think in about 20 years we will be there, these are things that
00:29:30owners of this type of...
00:29:35And we are moving very fast towards that, right? Notice that for this
00:29:38to happen, the bases have to be laid
00:29:42and the bases of AI are basically two,
00:29:46one, where do I store this amount of data and
00:29:49how do I move this amount of server, that is, energy, okay, and if you
00:29:53notice, many news stories have come out like, uh,
00:29:56Oracle
00:29:58already loses in sales with its ERP, but wins in the
00:30:02clouds that Antropic is giving it and Google is giving it, right?
00:30:06That is, for it to store in my servers and also
00:30:10the United States is talking about many more servers and energy and always
00:30:14the topic of energy, right? That is very much in vogue, why? Because if I am going to
00:30:17grow the general idea, those two bases have to be very
00:30:21robust. Our time is up, thank you,
00:30:24thank you for your time, we were talking with Gabriel Montiel,
00:30:28an AI expert in this program of your questions, until next
00:30:31Sunday.