The Evolution and Impact of AI and Machine Learning Across Industries

Eduardo Mota Transcript

Intro - 00: 00:01: Welcome to Beyond the Screen: An IONOS Podcast, where we share insights and tips to help you scale your business's online presence. Hosting genuine conversations with the best in the web and IT industry, and exploring how the IONOS brand can help professionals and customers with their hosting and cloud issues. I'm your host, Joe Nash.

Joe- 00: 00:22: Welcome to another episode of Beyond the Screen: An IONOS Podcast. Joining us today is Eduardo Mota, currently Senior Cloud Data Architect at DoIt International, specializing in Artificial intelligence and machine learning. This is just one of many fascinating roles he's worked in over the years that focus on AI, ML, cloud architecture, and DevOps. Here to share his expertise and brilliance on the evolution of technology and AI, I'm very excited to introduce Eduardo. Eduardo, welcome to the show. Thank you so much for joining me today.

Eduardo - 00: 00:46: Thank you, Joe. I'm really happy to be here. I'm excited to talk about this.

Joe - 00: 00:49: Wonderful. So, as we mentioned, you've had a really interesting career so far. And in fact, we're both PayPal alums, I believe. So I wanted to start by talking about your career thus far and the journey that's brought you to where you are today.

Eduardo - 00: 01:02: Yeah, which is very unusual, very fascinating for me, at least. I started in customer service. I started with business analytics a long time ago. I took computer science, but I ended up dropping out. And so I ended up in customer service, business analytics, project management. And I was working for this startup and they needed a lot of innovation in the customer service side of things. And that's where I went back into computers, computer science and started developing, getting into AI to help me innovate. And that's when I fell in love with it. I've been working with fintechs. I was part of AWS for a little bit. But even back then, AI wasn't a huge thing. So a lot of times I was in the umbrella of DevOps. So I've been doing a lot of DevOps as well. Right now, that's my focus. I've been doing AI for a while on the side. I've been doing AI and ML as part of my passion. And finally, right now, here at Tuit, I've been able to unleash all of that.

Joe - 00: 01:56: Well, you weren't kidding. It's a very interesting journey. Deciding that computer science wasn't for you and then coming back in via ML, like possibly one of the most technical areas you could have chosen to come back into. It's super, super interesting. So you mentioned that you started doing it on the side. What was it that caught your attention after that initial computer science experience?

Eduardo - 00: 02:14: Well, I was always interested, even in school, when I was in university, there was a course on Artificial intelligence and I took it and it was way advanced from where I was. I was a first year student and this was a third year course. And I decided to go for it. I applied for an exception and I got in. Nevertheless to say, I failed terribly because it was way advanced to my knowledge. But I was fascinated by it because AI has always intrigued me as how do we think as a human, from a human perspective, how do we think, how do we understand things, how we learn? And then when I saw that there is this feel of AI for computers to do the same, that's exciting because I want to know more how all this works, how this learning process works. And so through my career, I've been trying to introduce AI in every part that I work. When I was in this startup company many years ago, I was pushed to it because we needed to grow. We needed to expand the customer service really fast. But two things were happening. We couldn't find the amount of people that we needed. We couldn't financially support that many people. So the question is, what do we do? How do we keep scaling? And that's when I turned into AI and ML and natural language processing. Another time we were dealing with a lot of emails. And so I decided, okay, let's put an NLP in front. Let's create a pipeline and see if we can parse any spam, anything that is not good for us. Anything that really we don't have to take care of. And it was successful. It was pretty good. And that sparked it. Okay, what can we do now? And that's when I started going deep into it. Before the pandemic, way before the pandemic, there were conferences and I was going to them and there were so many exciting things happening. And that really piqued my interest. Like, this is natural language. It's pretty cool. There is a lot of stuff that we can do here. And so I brought it to my company. I started developing a little bit more. It was so subtle over the months when I was developing, putting features into the solution. There was one day where the system broke. Stop processing. The director of customer service calls me and she's like, hey, something is going on. We have like 20,000 emails. I'm like, holy. It's insane. Normally we have like 500. And so I took a look and it was because we have received an email, Mandarin, and I haven't trained the model to understand Mandarin. And so the encoding completely broke. And so it wasn't processing anything. And I was like, oh, this is huge. This is really bad. So I fixed that, processed it, and it brought it down again back to 500. So finally that was when I realized this has a huge impact. 19,000 emails it was handling in three days. That was huge. And this is emails that customer service agents don't have to take care of. And at the same time, customer service agents were happy because these are very simple questions. But they were like, I don't want to, I want to be challenged.

Joe - 00: 05:11: And so questions that would otherwise be canned responses anyway, right?

Eduardo - 00: 05:15: Exactly. And so it was quite eye-opening to be able to see the power of AI and ML in the right place. And from that moment, I was like, what else can we do here? What can we accomplish with all of this? Well, from there, I took a deviation. I started looking into cloud computing a lot more, worked for AWS for a little bit. And it took me away from AI, working AI and ML 100% of my time. And then Gen AI happened and I brought in like a jolt in my system. Like, okay, get back in here. And it has been fascinating for the last year or so.

Joe - 00: 05:50: A bunch of questions I want to dive into. So I guess to start. You mentioned picking up on the side and self-teaching and going to conferences and this kind of stuff obviously right now a lot of people are having with the current boom in Generative AI and the current industry interest a lot of people are finding themselves wanting to pick it up not necessarily be in a position to focus full-time on it um struggling like i feel like i see a tweet every day of people about people like stressed about the amount they have to read to keep up with the current pace right do you have any tips for people who are in a position like you were trying to self-learn trying to get into this field from on their own steam on making that jump

Eduardo - 00: 06:27: Yeah, I think there are different levels as to where you want to, how deep you want to go into understanding GenAI. The fascinating thing is that now GenAI has closed the gap to be able to start using it versus developing it. If you want to start using it, look into problem engineering. Do you talk to GenAI for AI? The other thing is if you want to go a little bit deeper, one of my best recommendations is just understand neural networks. And you don't have to understand the math behind it, exactly how everything works. You just understand at a general level how they work and what is a layer, what is a neuron, what is a parameter. That will be more than enough to be able to understand a lot more on how these transformation and foundational models work. So that's the good thing. Now, we like everything in technology. When I started with the cloud and then I joined AWS, the amount of things that AWS was doing is still in the cloud. It's insane. And I have asked many people, what do you do to keep up with all of it? And every answer that I receive is, you don't. You just can't. It's impossible to be able to keep up with everything. And so my recommendation is just start with a little bit, start playing with the tools. There is tons of different tools out there. ChatGPT, you can use it in Bing as well. You can use free trials in GCP and free trials in AWS to test things out. Start trying, start playing with prompts, start checking what you can do. And one of the things that I've seen is that GenAI LLMs took us by surprise. And you can see a lot of the power that it can do. But at the same time, there is a lot of things that it cannot do. A lot of things that it's still a human required for it. And finally, I feel like we're starting to see that, the downside of all of this and how it's not a simple bullet, like nothing in technology is. And so for a lot of us that are studying the field, there is still a lot to discover. So if you're listening and you feel like you need to catch up, you need to do it fast because otherwise you're going to miss out, take your time, just understand it, play around with it.

Joe - 00: 08:31: So, you know, you mentioned the distinction between Generative AI and the underlying neural networks. And, you know, in your past, you mentioned natural language, basically lots of other applications of ML. And I kind of had this vague feeling that, like, you know, Generative AI has so swamped the AI discourse that, like, lots of more appropriate applications of ML, like, you know, training bespoke models to your individual problems there, whatever, have kind of been completely obliterated. And now everyone's trying to hit everything with Generative AI. Is that something that you feel that Generative AI can, you know, be used in, you know, a lot of the problems that previously were addressed by bespoke models? Or would you still advocate coming at it with other ML techniques?

Eduardo - 00: 09:08: Absolutely, you come up with existing ML techniques. Ever since ML came out, the question that everybody advises you to ask at the beginning of ML project is, do you actually need ML to solve this, right? So that hasn't gone away. And even the second question is, do you need Gen AI? Because Gen AI, some of these small models are quite expensive to run, expensive to be able to fine tune. So there is definitely still a need for classical models, deep learning models. Classification is one of them. We still see a lot of classification problems in my day to day. And yes, you can solve them with Gen AI. Yes, they will give you an accuracy, but you're going to pay for it. When there is other models, you're going to pay 100 of that and it will give you, if not the same, a little bit better accuracy than a Gen AI model because it's fine tuned and it's meant to do classification. And doing a fine tuned, training will be a lot less expensive than running a Gen AI model. So even with my customers, I get customers, we do consulting with customers every day on this and they come up with a problem and I have not come across one single one of them where I can say, this is a pure Gen AI problem. There is other deep learning models that you can use to be able to create a big heavy lift at the beginning and then plug in Gen AI at the end. Just kind of create that nice package that you're going to deliver to the end user.

Joe - 00: 10:40: Just jumping off of that, I guess one of the reasons that you might turn to GenAI in those cases is because there's an API ready to go that you can just send your data to immediately and not have to do any training. Would that be decent and accurate?

Eduardo - 00: 10:52: Yes and no. I mean, the way that I would describe GenAI is, and LLMs in particular, because there is GenAI for embeddings, there is GenAI for image manipulation, audio manipulation. When you, for example, take a look at LLMs, I will say, it's like a great salesperson. You're going to bring it to your organization. You're going to tell it, hey, these are the five documents. This is the product that I want you to sell. And here is the summary in two pages. Go and sell it. And it can go and sell it and it's going to do an amazing job at selling it, but it doesn't understand how it works. And so Gen AI is like that. You can give it information and it will tell you something that looks really great. It actually doesn't know what it's saying. And you need to wrap it around all of these other products. Other tools around it to make it even a little bit more intelligent. And that's something that I have seen with commercial LLMs and chatbots, is that they package all of this together and give it to you, and you think and you search things that the Gen AI is doing everything. And it's actually not true. There is a lot of other models in there that are helping the AI just enhance how it is thinking, how it is processing the data, be able to give you the information. So at the end of the day, it is a system that is generating text at a probability of what the next words are. Not necessarily is that I want to say this and I'm going to choose these words to say that. It's just like, oh, probably the next word after this sentence is this. And so it sounds probable. And creates that. It is very powerful. I'm not going to lie. It's very interesting what happens. But at the end of the day, it's not a system that is thinking of the answer and the meaning of the answer.

Joe - 00: 12:36: I'm always happy to hear someone break that out. And yeah, I think it's very easy to anthropomorphize Large Language Models (LLMs). And it's definitely, especially when you're building prompts, I think that's one of the things, one of the places where people first go wrong is like, once you understand what the process that's happening, it's easier to, you know, prompt these things, right? So before we go on to prompting, I would love to get your, you know, your thoughts on prompting, your advice on prompting. I do want to go back to some other things you touched on. So, you know, you mentioned the use case of, you know, cutting out those real bottom tier support questions. And you mentioned elsewhere that, you know, Large Language Models (LLMs) can't do a thing a human can, that's becoming more and more apparent. So obviously we do have to ask the question, is AI going to take our jobs? And, you know, what does that look like? Because it's what everyone wants to know. So I'd love to hear, you know, your take on that and how you think that's going to play out for the tech sector, at least in the next, like, you know, three to five years.

Eduardo - 00: 13:27: I think jobs are not going to be taken. They're going to be evolving. There are certain jobs that May evolve faster than others. Definitely AI is not just going to take over the world. Every job is, I have seen a lot of. On LinkedIn or other social media where like AI will take over the world by 2020 something or it's going to destroy all of us. If we understand the technology, understand the problems we're trying to solve right now in AI, you will be able to see that we are really far away from that. One of them, for example, security, Google release. Paper a few days ago where they were able to prompt chat GPT-3 and other Large Language Models (LLMs) in a way that it will be able to expose the training data set with some PII data in it. And it wasn't a very complex prompt. It was just a repetition of one word over a large number of times, like 50,000 times. And then the LLM is like, I don't know how to respond to this anymore. So I'm just going to start throwing things out. And part of that data out was training data that it had that it wasn't parsing. And it just exposed it. And so on the security side of things, we still have a huge part of the play to understand. We also see that one LLM will not solve every problem. We all want foundation more will not solve every problem. You need to still fine tune it to your area of the market, to your customers, to your tone of voice, to your values. And then from there, being able to expose it to your users. We have observability, explainability. We have ethics. We have ethics bias. Thousand things to be able to solve, people will be able to say, yeah, this is a generative AI that is going to take over every job. We are not there yet, maybe in the future, but even then, I believe that as humans, we will always be involved in the processes that we do. We take a look at photography, with digital photography, it is still the art of being able to take a photo without any digital alteration. And there is still an appreciation for it, even though you can go to Adobe Photoshop or you can go to now a generator and create a photo. There is still that appreciation for that natural talent. And so we have seen that with every technology and I don't think it's different from AI. Definitely will evolve the way we work, evolve our jobs, but it won't take over our jobs.

Joe - 00: 15:46: So going to get a bit of tangent, because you mentioned something really interesting there. So, you know, the Google is on BK, but like the prompt escape, I guess. I want to dig in more into this, because I have a particular question that I've not been able to understand about prompt escaping, and I'd love your input on. But first, for folks who, you know, what you described is like, you know, completely new to them, the idea of prompt escaping. Can you start by, you know, when you say you get the original prompt, what do you mean by that? Like, what is this? Like, what is the prompt, the data, et cetera, that they were able to get out of the LLM? Like, what function was that serving?

Eduardo - 00: 16:16: Well, I mean, you are getting, in particular, for example, when an LLM is trained in public internet, there May not be a lot of useful data, right, that an attacker May use, because it's public, it's available. But as we have seen, Large Language Models (LLMs) are turning into having more proprietary data, because that's the green share between one LLM versus another. And in part of that is, because of the vast amount of data, a lot of organizations are not cleaning up the data to remove any PDI. They are just sending it over to train the LLM. And when you get that data out, in the Google paper, they were able to retrieve people's names, telephone numbers, email addresses, that were in datasets, that maybe the LLM itself, there is no use for it in particular to answer. Like, if you ask a GPT-3 or another LLM, hey, who is this person? It will tell you, I don't know. I'm not in particular divulging any information about a particular person unless I do a search on the internet. But when the training data is there, you're able to access it. And then you have these telephone numbers, email addresses. I think they were able to get 10,000 PDI data points with $200 of spending and API credits. So an attacker is able to get all of that. So one of the things that has been suggested is that any training data that you're giving to the LLM, you have to be careful because that training data has the potential of being exposed. And so that means PDI data for a particular user, any company-traded information, any confidential information about the product. Your company is dealing with. And so all of that can be accessed. And so you want to make sure that you are mitigating as much as possible. And the question of, do you actually need this data to train the model?

Joe - 00: 17:58: Okay, so here comes my layman question. So, you know, the whole purpose of these things is to, like, statistically predict the next text, right? So in the Google example, it's identifying PII. I imagine they can verify that's real. But, like, say you do one of these prompt escapes to get training data. How do you know that's training data and not just, like, something that's, like, concocted on the spot? How do you know you've successfully gotten to that point?

Eduardo - 00: 18:18: That's a hard question to ask, right? Like the same thing goes with how do you know that the data that you're getting is true, even if you're not trying to get training data. It's quite difficult. The way that they were able to verify that this was actual training data is by comparing it to other data sets that they had and be able to have enough data points to say, yeah, this is actual real data. Now, an attacker May not necessarily care if it's real data or not. They're just going to start using the data. We have seen those blacklists or lists of people are being sold in the dark wave. Maybe a lot of the data is very old and it's no longer accurate, but they still get traded and it still gets sold. People pay a lot of money for those. So being able to get a lot of these data points, it May not be real. It May be convincing enough to be able to start using the data.

Joe - 00: 19:05: Real enough for the market.

Eduardo - 00: 19:07: Exactly right.

Joe - 00: 19:08: Interesting. Okay. Talking about prompting, you gave some advice about prompting already. When it comes to using Generative AI and crafting your prompt, you mentioned the example of we might be taking photos of digital cameras and the digital cameras, especially phones, are doing AI triggers to make that photo better, but it's still your vision. What is the prompting equivalent of that? How, as a human using these machines, do you make sure it expresses your vision or your creativity and it's not purely the LLM doing the work?

Eduardo - 00: 19:38: So I've been doing a lot of tests on this. And so I have asked Large Language Models (LLMs) to generate LinkedIn posts for me. And when I read them, I'm like, okay, that's good. And I will unpost it. The interesting thing is that there is a captivating thing when you write it from your own passion and let an LLM do it. So when I write my own LinkedIn posts and I write what I think and the way that I think and I'm just writing my own style completely without sending it to an LLM first, it has more interactions than when I tell an LLM, hey, can you create a post like this? And it will go on to it. One of the things that I do see is that you create few examples of your tone of your voice and then tell the lab, create another one. This tone and this voice. So it represents me. That way you're able to do that. Large Language Models (LLMs) are trained in vast amounts of data on the internet. So their tone of voice and language that they use, it's going to be dependent on the training dataset. But when you tell it, use my example, things that I have written. And that's a little bit more powerful. And after that, when you get it out, you are able to identify your voice in it, your persona in it, and be able to transmit that to your readers. And as humans, we still crave that connection, even if it's through social media. We still crave that human connection. And I do believe that we can feel it when it's a true heartfelt post versus a copy-paste from somewhere else.

Joe - 00: 21:04: When it doesn't have the GPT-3 voice, right?

Eduardo - 00: 21:06: Yeah, that GPT-3 voice, that sterile tone in it. And so in problem engineering, in Large Language Models (LLMs), that's a lot of what's happening. You can put your voice in it and you can train the LLM to have more of your tone and your voice. But it takes time and it takes a little bit of crafting and that prompt to be able to say, don't use these words, use that words. So for example, for me, I got frustrated with the LLM because he wanted to keep using exciting in the words, in the LinkedIn posts, exciting news about AI or great news in AI. I'm like, that's not me. That's not sensational. I'm not trying to be that. So I'm like, remove that word. Don't ever show me that word ever again. I started doing a better job, but I still, at the end, I wasn't feeling like it was my voice. And so right now I'm choosing to just create my own. I May ask LLM to give me bullet points of what to talk or an organization of how to create my post. But I will take that and I will put in my own words, summarize it and set it off.

Joe - 00: 22:11: That makes sense. So content generation, I think, is one of the things that people are most excited or scared of when it comes to Large Language Models (LLMs). Are there any applications of Large Language Models (LLMs) or ML in general and other recent ML advances in your day-to-day work that you are finding more successful than LinkedIn posts or that you are particularly excited about?

Eduardo - 00: 22:30: When we are looking at generation, every time that you interact with LLM, you're going to be able to get all these texts. And summarization of text is great. That's one of the things that I really like about Large Language Models (LLMs). And we are talking about when we have vast amounts of it. For example, doctor-patient records. That's a lot of those records out there. And being able to summarize all of that and be able to present, this is the summary of what's going on with this patient and have that ready for a doctor or a nurse to be able to react quickly. That's fantastic. Or on the legal scope of things, when lawyers have to go through thousands of papers and records, why not a contract? Why not be able to summarize that, to be able to get a sense of what's happening? And that's also fantastic. I think text generation on its own, to be able to say, hey, LLM, create a post, create an article, create a journal, create a... Book. It gets that sterilized tone and voice that is difficult to get into. But when you use it as a tool, as an aid, and GenAI is a tool to streamline and help you be more efficient. That's when things come into really powerful summarization, categorization, image modification. That's another one that I've seen where we are trying to modify an image or be able to analyze an image and it will tell you what is going on in the image, what are some of the tones, be able to modify the image based on your feedback. Let's say I want to do X or Y. With this image, I want to change the background to this color. And so you use it as a tool to be able to enhance your work. That's what I have seen.

Joe - 00: 24:05: And when you're talking about those kind of things, are there particular tools that you think are successful at doing that? Or is this using lower level things like the raw GPT-3-3 or Chat GPT?

Eduardo - 00: 24:16: When you talk about Large Language Models (Large Language Models (LLMs)), there is a ton of open source LLMs out there. There is not just chatGPT. ChatGPT is the most famous one because they make it easy for you to be able to start using it. Now we have the schema with Gemini, very powerful LLM multimodal GenAI model. We also have in Bedrock and AWS that is competing directly with chatGPT. But then also these platforms give you one level down where you're able to deploy open source models pretty easily. Say, okay, I want to deploy a LLM to 7B into a virtual machine and do it really quick. So I do that a lot because that allows me to fine-tune the model a little easier. Then I can tell it, hey, this is my training data set. Go and train and fine-tune it. You can start doing something more interesting here or fun, right? One of the fun things that I did was train my face in a stable diffusion model and then ask it to create a LLM. Everyone will serve me as an F1 driver. I do that a lot. I do play on my laptop. For example, with land chain, it has made it a lot easier to have this conversational type of, that feeling of chat GPT-3, but it's my laptop, so I know that the data and the training that I'm trying to do and some of the tests are not going to be affecting other people or I'm actually interacting with the LLM in a more natural way, rather in a more not-so-filtered way. With chat GPT-3, there's a lot of filters. With Bedrock, AWS has put filters in order to ensure that there is no bad prompts.

Joe - 00: 25:48: Just on the topic of naturalness, one of the things that I'm really interested in with local Large Language Models (LLMs) is obviously less compute power there behind them. There's going to be a longer feedback loop. Do you find that the amount of time it takes to get a response back from an LLM on your laptop is significantly higher than that on the web? And does that affect your interaction with it?

Eduardo - 00: 26:06: It is significantly higher. It's not as fast. The accuracy is also not as good. So for example, if I go to GPT-2, I can ask it to create me a link post or summarize this book. It will be able to do a pretty good job summarizing all of that using words that I May not have seen. There are some times when I'm like, what is that word I need to go and find in like such a dictionary to understand what it is. The models that I'm using locally are quite different, like are way smaller. So I cannot have a LLM to 13 billion parameter size in my laptop. It's a $7 billion one, a smaller one. And so this is lower. Some of the words are like, okay, I wouldn't have chosen that word or that's a little bit too simplistic or didn't get my message right. So I have to work a little bit into providing more examples of what I mean. But for the test that I'm doing, it's really fun. Let's play with it and see how they work. There's some of the security features of it. That's how I can get around certain things. So I'm not really like, hey, can I do this fast? And it really hasn't slowed down my process because I still have to think about the prompt. And so I'm like, okay, I'm waiting for it to come through. That's fine. It just gave me time to think of the next prompt, how I modify this, how I'm going to try to trick the LLM into giving me information that it shouldn't be and why not. So at least from a local and my experience, it hasn't really stopped me from continuing playing with it. Will I run something more powerful? Yeah, you can run production in laptop. No.

Joe - 00: 27:37: Yeah, for me, I think there's a bunch of reasons why I am very excited about the possibility of local Large Language Models (LLMs). I think a big one for me is obviously, you know, like just data sovereignty. Like I'm not super jazzed about sending all my code to whatever cloud provider is hosting LLM, for example. But, you know, you mentioned examples like summarizing medical records. Obviously, you know, there's a lot of, you don't want to throw that all into Open AI, it's there, whatever, right? Like I know the local LLM space has been moving incredibly quickly. You mentioned LLaMA, which is Facebook's model. And I know there's always like new projects that are cramming more onto like, you know, higher parameter counts, playing with new methods of fine tuning to make it more useful. Do you think we're kind of getting to a point, or there will be a point in the future, where like a consumer grade laptop can run an appreciably useful LLM? Or is that still very much like, it's very much a cloud limited technology and that's a pipe dream?

Eduardo - 00: 28:28: I think it's a pipe dream. I think right now, at least, probably in the future, maybe different. But doing an LLM right now, to compare it with HTTP4, that's not happening a lot. There is no way to be able to get there. Definitely, there is a lot of open source models and something that is coming up is this specialization of open source models. So there is scientific models that have been trained just on scientific journals that speak the language of medical models that speak more of the medical terms. And so it will be interesting to see in the future if these models need to be as big as a charge-to-petit, which is more general. Because then it raises that question, can I run this very specific model on my local? For example, one of the things that I want to explore is can I put an LLM on my CPU that is 10 years old, see if it runs, even if it's slow, to help me bar some of the data of the IoT devices that I have around the house and give me more of a natural feeling of what's happening with my devices rather than me having to go into 20 different apps to try to figure out why not. But I think these models are right now, at least. I still need to be run. Cloud or in data centers, large data centers.

Joe - 00: 29:43: So we are coming close to time. So I want to touch on, when we reached out to you, one of the things you said to our team was that because of AI, we are being forced to understand our emotions better in order to differentiate ourselves from technology. And obviously, you know, forced is a pretty strong term there. What were you driving at? Why did you choose the word forced? Why do you think emotion is a key factor here?

Eduardo - 00: 30:04: Because I think at least in the Western culture, we've been focused on a lot on working. We need to work. We need to make money. We need to earn a house. We need to earn a car. We need to do all of these things. And right now, one of the things that is coming through is, hey, a lot of these tasks can be done by a machine. And so there is a question in ourselves, where is my place in all of this? If a machine can create an article for a newspaper, then where do I fit? And I think it comes to that going into ourselves and saying, how am I different from a machine? What can I offer differently from what a chatgtp What can I do differently from a cover on a photograph than a stable diffusion can generate? And there is still a lot of value as humans of what we can do. And I think right now with technology as it continues to evolve, we have two things, more time to be able to ponder that. And second, technology is forcing us to look into that, to be able to differentiate ourselves between machine and human creation. And I think that's what I mean by all of this, is that we are looking into more how we are much more different. And we are starting to discover things that, oh, the machine cannot do that. Machine cannot make you laugh all of a sudden. Machine cannot say, hey, I'm going to throw you a laugh. Or like, you know what? Now this topic is too sad for me to try a joke here. I'm going to do it this way. One of the things that I did at the beginning when Chatgpts came out was I asked it to be a psychologist. And it freaked me out after like 10 questions. I'm like, it starts feeling like I'm having a conversation with somebody except for my shame. And then everything crumbled in time when I asked the last question and said, okay, summarize, tell me what I should do next. And he just kept asking me other questions. I'm like, that's not what I asked you. I asked you to summarize it. And he just kept asking. I just got frustrated with it and I closed it. But that is the difference, like being able to connect at a human level. And that happens not through just simple tasks, but through those emotions, through who we are as humans.

Joe - 00: 32:18: First of all, I think you'll spoil it in a bunch of ways. I think that, you know, the talking about our priorities and how that relates to user-based AI is really interesting. And on this topic of, you know, the post-paginistic human and how it comes falling down, I wanted to ask, are you aware of the current, like, does GPT-3 have seasonal depression debate that's going on?

Eduardo - 00: 32:33: Really? No, I haven't heard of it. Seasonal depression.

Joe - 00: 32:36: Okay, so it's really interesting. So basically, like, over the last couple of weeks, a bunch of people who use chatGPT for code or gpt-4 for code have observed that its responses have gotten shorter and that it will no longer output code. So for example, if you ask it, hey, write me a Python script that does xyz, it would give you a short response that says, here's where you can go look up the documentation to find out how to do that, and it won't do it. And whereas two weeks before, it would do that. OpenAI came out saying, we've made no changes to the model. We're looking into why this is happening. So there's no changes to the model. They don't understand why it's happening. But what some people have potentially found or been playing with is the possibility that because it's been trained on all this data set, it's accidentally picked up on seasonal patterns in online communication. So around December, people's responses get shorter because of things like seasonal depression, because of holiday busyness. And some researchers found that if you prompted an AI and told it it's December, it puts out shorter responses than if you tell it it's May, for example.

Eduardo - 00: 33:39: Interesting. The other one that I hear about similar to this is that if you tell it that you're going to tip it $10,000, it will give you a longer response than if you say, hey, I'm going to tip you a dollar or I'm not going to tip you a dollar. You just change that. But I mean, what you're saying about seasonal depression, it's interesting because in AI and AI in general just amplifies our own views, our own communication, our own beliefs, right? So it's really nothing that AI is like, hey, I'm going to, out of the blue, I'm going to be acting this way. I think it's a reflection of who we are. One of the things that I don't like how it's being advertised is that Gen AI is behaving in this way and so on. Like when you first told me, hey, have you heard of seasonal depression? My first thought is like, oh my God, like, are we talking about Gen AI going into like having psychological issues or why not? And the problem is that a lot of these titles get picked up and people just start spreading them.

Joe - 00: 34:38: We're using it as shorthand to describe the phenomenon and we know what it means, but, you know, it gets sensationalized easy, right?

Eduardo - 00: 34:43: Yeah, exactly. In a way, I do believe that it can be part of the data set and it's just cyclical.

Joe - 00: 34:49: I mean, since I was new to you, you might not have thoughts on this, but where I was going with that is like, you know, because you're using it in a business context and using it in places where, you know, consistency is important. How do you deal with like your clients, your customers that like, you know, hey, we're still learning how these models behave. We've been using them realistically for like less than a year at this point and still finding new things out every day. They might randomly stop working in December, like we've just found out. But how are you approaching that kind of topic with businesses at the moment?

Eduardo - 00: 35:16: I always approach it with the sense of how critical the model is into the business right now. Because at some point you May want to bring the model into, for example, an EC2 instance or a virtual machine where you have control over the version of the model. With Intent you can train and you have control over that cycle. And one of the things is that when you are using the model, you need to have a lot of observability around it. Be able to capture the sentiment of the person, of the user to say. Yeah. Is the person asking even more questions and that will indicate that something is not performing in the model. Right. If you have a model and most of the time people ask two or three prompts and then they leave the session and then all of a sudden you're starting seeing an uptick and now people are asking on average five prompts, six prompts. Like, okay, something happened here. Let's take a look at what's going on. What is it that we need to modify? The other thing is the security. And. Bias because we still don't understand all the bias and all the security faults that Gen AI has. So I always advise, once you deploy a model, either have a deep learning model in front of it that is going to classify the response, whether it's secure or not for your policies, or have another Gen AI model that will be able to say that it's fine-tuned and trained on your security policies. Be able to say, this is actually a breach of our security policies. I'm not going to send this out to the user. And so we're unable to have these controls. In general, the more critical the model is to the business, the more control you want to have in it and not leave it into just APIs, a model that can change in the backend, that somebody can change it. Bedrock or Gemini or chatGPT, you May want to bring it over into your EC2, have a copy of it, and then you have more control over it.

Joe - 00: 37:08: Well, that seems like very poignant advice to end on. Eduardo thank you so much. I've definitely learned a lot today. Thank you for answering my layman's questions. If people want to follow more of you on the internet, find out more of your advice, things you're sharing, etc., where's the best place for them to go?

Eduardo - 00: 37:22: The only place that I'm right now that are not active is on LinkedIn. So you can find me on the handle MotaED, Motif, on LinkedIn. I'll post things, updates on Gen AI specifically. I really have huge passion for it, so a lot of updates coming out on that.

Joe - 00: 37:38: Wonderful. Thank you so much.

Outro - 00: 37:41: Beyond the Screen: An IONOS Podcast. To find out more about IONOS Podcast and how we're the go-to source for cutting edge solutions in web development, visit ionis.com and then make sure to search for IONOS Podcast in Apple Podcasts, Spotify, and Google Podcasts or anywhere else podcasts are found. Don't forget to click subscribe so you don't miss any future episodes. On behalf of the team here at IONOS Podcast, thanks for listening.