EP12: AI's Next Wave – OpenAI's Operator, Nvidia's Music Venture & Uber's Training Push

In this episode, hosts Marlon Avery, Sekou Doumbouya, and Adrian Green explore the latest in artificial intelligence, starting with OpenAI's ambitious plans for computer-using agents—set to redefine user interaction by 2025. They discuss the potential implications and competitive landscape, touching on privacy concerns and automation.

Nvidia's AI Music Revolution

The conversation shifts to Nvidia's Fugato model, a groundbreaking entry into AI music generation. The hosts analyze its capabilities and potential impact on the music industry, drawing comparisons to other tech giants.

Uber's AI Training Expansion

Next, they tackle Uber's new division, Scale Solution, which connects companies with contractors for AI model training. The discussion delves into ethical implications and the challenges of fair compensation in a global gig economy.

Blue Sky's Data Privacy Concerns

The hosts examine Blue Sky's recent data privacy issues, highlighting the balance between open APIs and user trust.

Zoom's AI-First Rebranding

Finally, they analyze Zoom's bold rebranding as an AI-first platform, exploring its expansion beyond video conferencing and the competitive landscape with Google and Microsoft.

----------------------------------------------------

Welcome to AI With Friends, your weekly launchpad into the world of Artificial Intelligence. Hosted by Marlon Avery, a pioneer in GenAI innovation, alongside Adrian Green, VP of Engineering at LiveNation, and Sekou Doumbouya, Senior Staff Cloud Systems Engineer, this show is your go-to source for all things AI.

Our hosts bring diverse expertise—from AI strategy and tech innovation to industry leadership. Every week, they break down the latest AI trends, interview top experts, and simplify complex concepts for AI enthusiasts, entrepreneurs, and tech professionals alike.

Marlon, Adrian, and Sekou combine their unique perspectives, whether it’s Marlon’s collaborations with tech giants, Adrian’s leadership in global entertainment engineering, or Sekou’s cloud systems expertise. Together, they make AI insights accessible, actionable, and exciting.

Tune in live on Twitch & YouTube every Wednesday at 6:00 PM PST/9:00 PM ET, or catch us on all major podcast platforms.

Here’s what you’ll get:

Cutting-edge insights from AI leaders
Real-world applications of AI technology
A vibrant community of forward-thinkers
If you're ready to stay ahead of AI trends or spark your next big idea, join us each week for an hour of engaging, thought-provoking content.

Subscribe now and become part of the future of AI with AI With Friends!

Note: The views and opinions expressed in this podcast are solely those of the speaker and do not reflect the official policy or position of any current or former employers.

Transcript

Marlon Avery: Are you guys ready for Thanksgiving? Because I am extremely ready. 

Sekou Doumbouya: Oh, yeah. 

Marlon Avery: Hey, wait. No. I have one controversial question. Is there a such thing of eating breakfast on Thanksgiving? 

Sekou Doumbouya: I don't know. 

Adrian Green: You know, of eating breakfast on Thanksgiving, but. Wait, on Thanksgiving Day? 

Sekou Doumbouya: Thanksgiving? 

Adrian Green: Yeah. 

Marlon Avery: On Thanksgiving Day? 

Adrian Green: Absolutely not. Absolutely not. Are you kidding? Like, I mean, I think that you can probably put some science behind it. You know what I mean? So if you eat enough to, like, perfectly. You know what I mean, really get your appetite going just to kind of prime it. Because if you. If you're too hungry, then you can't eat. And I've definitely done that. I definitely have done that. So what I'll do is I'll maybe do a like three hour before snack or something just to lay the ground. 

Marlon Avery: Just. 

Adrian Green: Just to lay the foundation for. But just how I'm going to destroy my stomach, you know? 

Sekou Doumbouya: You know, my strategy is I usually just. I just stealthily walk into the kitchen when no one is looking and, like, grab the scraps of these and then, like, the cheese and stuff that's laying around. Like, that's my strategy. 

Marlon Avery: That's awesome. 

Sekou Doumbouya: I have, like. It's almost like a. It's a pre Thanksgiving, you know, grazing. Grazing. There you go. 

Marlon Avery: Doing grazing. 

Sekou Doumbouya: That's exactly what I do. Every now and then I get caught and get kicked out of the kitchen. 

Marlon Avery: Yeah. 

Sekou Doumbouya: And, you know, that's how that goes. 

Adrian Green: Oh, yeah, that's on there. 

Sekou Doumbouya: Yeah. 

Marlon Avery: On my end. There's no such thing as breakfast. Breakfast doesn't exist on third on. On that Thursday. Yeah, it's. First of all, people eating breakfast on Thanksgiving can't be trusted. Let's start there. You know, you're going. You're going against everything, you know, this holiday has been built for. What I've started doing lately is I've started doing like a workout or like a walk or run like that the morning of Thanksgiving and everything, you know, to prepare, you know, for that day. 

Adrian Green: That's exactly what I'm gonna do tomorrow. I'm doing that. I'm actually gonna get out there and work out tomorrow. Hopefully. Hopefully. It's good. It's good weather. 

Marlon Avery: And then I watched. I'm gonna do that. And then I watched the. The movie 300 to get my mindset ready. 

Adrian Green: Your family's like that, man. Go le, man. It's like that around the. Chase the Spartans. Okay, then Bill was like that. 

Marlon Avery: Just get it easy. Get it. Get. Get it prepared and stuff there. Yeah, that's a lot. That's hilarious. 

Adrian Green: Yeah, man, I understand. 

Sekou Doumbouya: I Guess we better get started, right? 

Marlon Avery: Yeah, a little behind the scenes action here. All right. Hey. All right, guys. Welcome back to another EP of AI With Friends. If you just join in for us for the first time, guys, this is a podcast where a group of individuals, software engineers, builders, architects, doers, if you say of artificial intelligence, we come together every week and stuff here. Fun fact. We've been doing this conversation in private for like the last year. Plus we've been having conversations around some of the things that we see that's happening in the world of artificial intelligence. Sharing some knowledge, sharing some tools, building together, if you will. And so one day we decided to say, you know what? This may be beneficial for the rest of the world. And then Here we go, Ep. 12 of AI with friends. So, guys, my name is Marlon Avery. I am a geni doer, builder, architect, garbage man, sir, if you will. Principal. I'm gonna let the fellas introduce themselves, man. Seku. Yeah, yeah, yeah. 

Sekou Doumbouya: Hey, everyone, I'm Saku Demboia. I'm a cloud engineer, old school, gray beard, UNIX administrator. Been doing at this thing for, I don't know, I just keep saying 22 and I'm realizing every year I'm saying 22 years, which means it's longer than that. But I'm just going to stop right there. Trying to keep aging myself. 

Marlon Avery: There you go. 

Adrian Green: Cool, cool. I'm Adrian. I'm. I'm kind of a. I'm approaching the OG level. I'm about to be over the OG hill soon and the. In the web development space. I've been doing this for about 12 years, but I'm a real legit OG just overall it. Because I go back way to Nobel with this. So, you know, it's for those who know, they know, you know, but that's me. 

Marlon Avery: Cool, cool, cool, cool. Yeah. So, guys, typically what we talk about and stuff here, like I said, we're builders, we're doers, we predicted a lot of things, we've built a lot of things, and now we get to share our knowledge from practitioners. All of us right here. You see, we're practitioners. We're in stuff in this space. And so typically we kind of like to discuss some of the trending topics and stuff that's happening in the world of artificial intelligence. We kind of get our point of view, you know, we tend to, we tend to agree sometimes. We tend to disagree sometimes, particularly especially things around like who's the best Batman anyway? And so. Oh, those type of things may get thrown in here. We may have the Pause alive. Also behind the scenes discussions come back to you guys and everything and then we'll be ready to kick off and everything, you know, but we, we like to kind of like dive and stuff in the world and everything and we've had a good eye on just kind of like where things are at it and so we like to share these things with professionals like yourself, hopefully with the, with the hope and everything that you'll be able to equip yourself as we kind of get start to build towards this AI centric world, say it's centric universe and stuff where we're going to live in. And so hopefully you'll be able to benefit it professionally, if not monetarily, if not just overall in life and everything. And so with that being said, let's kick it off. Here we go. OpenAI one of our favorites discussion actually you can't talk about anything with AI right now with OpenAI. OpenAI is saying they are moving beyond chatbots with AI agents that are going to take action. OpenAI is set to release groundbreaking AI technology in January 2025 codename operator which will move beyond chatbots to AI agents capable of taking independent actions on behalf of the user. These computer using agents will be able to execute tasks with minimal human intervention handling multi step process like booking, traveling or writing complex code. This marks a significant leap forward in the AI revolution transitioning from passive tools to active problem solver. While childbirths have been increasingly sophisticated, their practical applications are limited because they cannot perform actions in real world applications. OpenAI said the new agents aim to address the limitations by integrating with existing tools and workflows. This moves aligns with broader industry tend towards automation as companies like Anthropic and Google develop AI that can handle complex tasks on its own. For instance, Anthropic has now introduced their tools that automate creations everything with websites, spreadsheets, editing. While Google is integrating automations into Gemini AI platform and Apple is also reportedly working on an update. It will allow users to control apps and perform tasks simply with their voice. So Adrian man, let's start off with you. OpenAI said hey chatbots is yesterday's conversation. They're calling the computer using agents is to is a new conversation. What's your thoughts? 

Adrian Green: Well, they better keep up. I mean that's my first thought because I got to tell you the real oracle, the real canary in this whole agent conversation to me we talk about it every week is lang chain. You know that was when I realized that this was going to be something that's going to be really sticky was when I was using LangChain. Their movement into this just kind of just. I mean it makes me wonder where really Lang Chain is in this. You know, what they're up to, what they're doing in this, while this whole agent conversation is taking off without them. First and foremost, I don't doubt that OpenAI is going to do a great job with it. What I will say is that there are a lot of people who have been doing very well in the agent space or whatnot looking at you replit. So it's as far as the. It's going to be interesting to see with how that the adoption of that new format. I feel like OpenAI has really, you know, made its bread and butter off of just the chat interface. So will those same users use the new agentic type of features? I guess time will tell for sure, but I'm pretty sure I'm confident they have some good stuff up their sleeve as far as just out of the box agents capabilities that they're going to roll out with this, with this product. 

Sekou Doumbouya: Yeah, you know what, I was looking at this article and I've been reading Apple and what's competitive to open AI. I think it's anthropic right now. Right. A couple of months ago they released their computer use API. Yeah, I started building some things around that. I feel like this is directly in response to that in some way, which I don't know. OpenAI so far has some pretty good execution right now on some of these different things. So I do expect them to kind of push it up a notch. Anthropic having this feature already, meaning that they have a little bit of a head start. But like it's no, it's not like it's no like good user interface. To do this you have to like wrap it in something in order to like get it to, to in order to like take advantage of it. It's not like there's a, a desktop app that has this functionality where you can, you do the computer use. So no, OpenAI has a pretty good app. Like I use it like, you know, all the time. The fact that I am a little, you know, I'm a little worried about letting OpenAI rummage through my computer, seeing all the things that it sees and training its model based off of that because, you know, nothing's. Nothing's free here, right? No. Delicious. So I'm a little worried about that. But you know, I think this is an interesting, like it's an interesting interface to move from just the be able to chat and like, have this chat bot interaction to, like, almost have someone working on the computer next to you. Feels quite interesting. I just want to see the delivery. I would love to see, like, the demo of some of this to see what this really means. And sadly enough, it's a little departure from OpenAI, right. Because last time they released some of their big offerings, they said, hey, we're doing this thing. Here it is. At the same time where they talked about it, it was available for you to do something with it. How they're moving into this world where they're telegraphing to the future. 

Marlon Avery: It's like. Like someone like sauna. 

Sekou Doumbouya: Yeah, exactly. Exactly. So I don't know if that's like a change in their. On their strategy or. 

Adrian Green: Or what, but you think it has to do with. Or is it. Is it related to the benchmarks? The 40 benchmarks maybe not hitting where they want to. So maybe like, this is going to you. 

Marlon Avery: Did you. 

Adrian Green: You think that maybe that fact ramped up the effort on the agentic side? 

Sekou Doumbouya: I think so. I think so. Like, what I'm seeing, not like, I'm not saying, like, you know, people raving around about a new LLM and how it's like, able to achieve something new that hasn't been achieved before, but I've been seeing tools or reuses, reusage of the same thing in different places. Right. So. Which is kind of cool. I think we need this moment right now. We need the moment where we're actually going through retooling our. How we're presenting this to our users, building tools that kind of build all. That's a sign of maturity that we're trying to bring into it rather than just feature, feature, feature, feature. It feels like it's a good thing. But, you know, I do miss it. I miss, like, I feel like, you know, the days where we're just, you know, you know, new things are being pushed out and we're like, excited like, oh, hey, look at this drop just happened. Yeah, that's a sign that that's changing a little bit, right? 

Adrian Green: Yeah, it's. It's less a hobbyist activity, you know? 

Sekou Doumbouya: Exactly. So I don't know. What do you think, Marlon? 

Marlon Avery: Yeah, I think. I think I'm gonna start off with what Adrian said or like, do you think this is. Respond to the benchmark, you know, measure? I don't think so. I think this is simply to kind of like, show. Show competitive edge in the market and also kind of keep the consumer interested or on their toes, you know, if you will. So Apple does the same thing. You know, we get the whole, we get the whole unveiling, you know, on demo day and then the release is three months later, you know, or, you know, it's coming soon, you know, so it's just like, hey, we're building this. We're kind of like, you know, putting our chest out, if you will. And then, you know, we'll keep iterating and stuff on, you know, behind the scenes and stuff like that. And then secondly, computer using agents is very, is very interesting. I think this takes us into a world of automation that we couldn't imagine. And so one of the things I see so many opportunities of here is now I see a world where the agents would be the vehicles to go to point A to point B, but you still need passengers, luggage, fuel to get it there. Meaning passengers may be your passwords, fuel may be the data and stuff that you need to give it. Also too, you need to give it access to your debit, credit card information. Now what happens with that? Do you simply, do you embed this into your agent? Second question behind it, is this agent, is this user owned? Is this simply an OpenAI, you know, IP, I mean all of us OpenAI IP and everything. But is this a, is this bad? Just am I just going to trust all my data to be within OpenAI or would I be able to connect my one password, you know, to my agent, you know, will I be connected and stuff there? I mean, I think right now we don't know. And I think some of those, those questions will, will really determine how well this transition happens for your everyday consumer. And also too, I don't see the everyday consumer probably integrating this into their world in problems of Q3 at the latest. 

Adrian Green: You know, short of easy integration, you know, we're talking about a desktop situation. We're talking about it running your computer. You bet it it. They. The. It's paramount that it's easily going to be able to integrate into your email, into your things like that, if that's what you know you're going to want. 

Sekou Doumbouya: Yeah, I'm just, I'm just waiting for someone to, well, I'm, I'm probably going to be that someone to run this inside of a virtual machine. Just because I don't trust it. Yeah, but you know, one day, who knows, it might break out, you know. 

Adrian Green: Or the next you buy will come with it. 

Marlon Avery: Sometimes you get, sometimes you gotta pop out and show. All right, Only on and our friends, when you get a Kendrick reference, talking about artificial intelligence only here. Now we gotta, we got, we gotta talk to our lawyers to figure out can we say the N word on this podcast. But yeah, yeah, I definitely see this, I definitely see this as movement. I see this as kind of like going place. I think, I think if I'm, if I'm being, if I'm looking at competitive landscape, I would argue, and this is maybe a little surprising, I would argue if they execute right, that's a big if they execute right. I would argue Google has the hand up in this type of area computer using agents because they already have so much more of the consumer data already. They already have the credit card information, they already have passwords and everything. They already, they already have Google flights. You know, they already have the, to be able to book hotels, you know, stuff all these things like that. They already have the web browser aspect. And Google, Google I will the web browser for now. So that, that part. And that's why I said, that's why I say if they execute this right, because Google has all. They got a, they got a history of being first to research, last to consumer. And so it's just so many different times where they're just there first and just drop the ball, you know, something along the way. So if they get right, I think Google has a leg up in the computer using agents competitive landscape. Oh yeah, it's. 

Adrian Green: It's so crazy. I'm sorry, go ahead, Sekou. 

Sekou Doumbouya: No, no, no, I'm just agreeing. I just like, I'm like, I'm just thinking about it. You know, they have, you know they have the Chrome os, right? They have that. They have, they, they have ways to like do deeper integrations than what OpenAI can do. They have a web browser. Well, unless the DOJ takes it away from them. I don't do j. The fcc. 

Adrian Green: Fcc maybe. 

Sekou Doumbouya: Yeah. FC takes it away from them. 

Marlon Avery: Wait, was it Chrome that was going to break up? I thought it was something else. 

Sekou Doumbouya: Yeah, they're talking about taking Chrome away, but yeah, I feel like you're right. I feel like there's some potential that I think Google can actually take advantage of in here with Gemini. I don't know, they're moving pretty fast. I'm like every now I look over, every time I look over to Gemini, I'm realizing, oh, there's a lot more there than what I saw before, but it's just not as advertised as well. 

Adrian Green: I feel like it's weird because we've been kind of circling the drain about this conversation with Google. For years. And that is that there are another Marvel reference incoming. But like, they're the Thanos that's kind of sitting back in the tech world where, you know, everyone knows they have the most data. And if this is a data economy, any, you know, the knee, you know, the kind of like reactionary responses like, okay, this is a data economy. This person, they have all the coins here, you know. 

Marlon Avery: Okay, so watch this. I see your Thanos reference and I raise you if Google is Thanos. Since I'm in this area, I'm going to say Apple Store. Apple Store. Because Apple has consistently behind the scenes. Wait till everybody kind of build what they're going to do and everything. They come behind and they include it into the Apple ecosystem and then boom, that becomes the main thing. You know, they did that. They did that with a wallet. You know, they've done that with several different products and everything. And so, yeah, I could definitely see, you know, if, again, if they get it right, you know, if Google does it right, I could definitely see them being Thanos. But I can see Apple coming through and just. 

Adrian Green: That's the thing. Thanos. It's when Thanos decides that he. That the percentage of the population needs to be cut down, you know, and that's what that. But I think it's like the looming threat that, like, it. It's always a thing. We know Google can just do this. 

Sekou Doumbouya: Just, you know, there's. There's a they. 

Marlon Avery: They. 

Sekou Doumbouya: I'm pretty sure there's. There's caution on their side because, you know, they don't want to seem like a monopoly because they hold all these pieces together. You can start connecting these dots and say, oh, we own this piece. You own this piece. Oh, you can decide if a market exists or not whenever you want. Oh, wait, that's a monopoly. Like, let's do something at. Bye bye. So I think there's a possession that they want to protect themselves in some way. Even if they had, if they wanted to do something like that and utilize all the different verticals they have, I. 

Marlon Avery: Think that's a really good. Wait, hold on. I'm sorry, are you talking about Google or Apple? 

Sekou Doumbouya: Google. Okay, I think that would be a little bit challenging. But my question is if Apple Store and what you call it. 

Marlon Avery: Google. Thanos. 

Sekou Doumbouya: Google Thanos, does that make Xai Loki? 

Adrian Green: Oh, totally Loki. I mean, just. Of course. 

Sekou Doumbouya: No, Loki is a reason why the Avengers were started. I'm just saying. Yeah, he actually did start the Avengers. You know, he got. He made them all. 

Marlon Avery: I see. Really I see where you're going with this. 

Sekou Doumbouya: Take down the one person so that. 

Marlon Avery: I see where you going with this. 

Sekou Doumbouya: You know, I'm just saying. 

Marlon Avery: Okay, then who's Anthropic? 

Sekou Doumbouya: Who's Anthropic? 

Adrian Green: Who is Anthropic? Anthropic, I'm gonna say, is Black Panther. I'm gonna say Anthropic is Black Panther. Because meanwhile, while this conversation is going on, Anthropic is. Is. They're. They're having the agent conversations, and they're also talking to Palantir. You know, I mean, they're also getting in cahoots with, like, the more, you know, Wakanda, like, forces. I guess that's why I call them. I'd say that they're. They're definitely Black Panther. 

Marlon Avery: So hold on. 

Sekou Doumbouya: We got. 

Marlon Avery: We got. We gotta be careful here, because if. 

Sekou Doumbouya: If you. 

Marlon Avery: If you're saying Anthropic is Black Panther, that also admit. You're admitting that Ant Anthropic has the more advanced technology. 

Adrian Green: Well, in the comic books, they sold the Vibranium. You know, it's. It's really comp. 

Marlon Avery: Good in. 

Adrian Green: Not the most researched comparison, I'll say that, but it's. But first in that clean the mind when it comes because I. I find. I think that, like, they're. They're kind of under the radar in a lot of ways still, and that, like, I think that the big players still snatch the headlines when it comes to AI over something that Anthropic's doing. 

Sekou Doumbouya: So, yeah, you know what? The thing is, when I use. I use Claude and any of the tools from. From Ann, from Anthropic, they always seem very polished, but I'm always using it from, like, the API side or using software that integrates with it. Like, I never use the actual. Like there. The web, little app or the web on the website. Like, never use that at all. Not there's something bad about it, but, like, I've already, like, got muscle memory at this point in time. Like, when it comes to, like, using chat opening. 

Marlon Avery: Yeah, yeah, same. Yeah, very much the same. Okay, cool, cool, cool. All right, cool. Here we go. Moving forward. So Nvidia has now entered the AI music arena with a platform called Fugato. Nvidia, the leading computer chip manufacturer, has launched Fugato, an AI music model designed for music generation and audio editing. It says Fogado is told to be a swift army knife for sound, capable of creating music from text or audio prompts and editing existing audio with remarkable speed and precision. One of Fugato's key features is the ability to manipulate audio elements, such as adding or removing instruments while changing a voice, accent or emotion, and even making instruments intimate of animal sounds. This level of controls open up a new possibility for producers, composers and musicians in video claims. Fugato is trained on a vast data set of audio samples compiled over a year, ensuring high quality audio output and compliance with the copyright laws. The company emphasizes that the model's emergent properties, which arise from the interaction of its various training abilities, enables its combined free form instructions. While Nvidia envisions numerous use cases for Vergato, including scoring visual music media, editing specific parts of music and manipulating voices, there is one catch. It's currently an internal research project and not available to the public. So this move signals Nvidia's ambition to challenge existing AI music models and its potential reshape of the music industry. Adrian, we're gonna start with you again. You know, you. You have the entertainment background specifically, you know, in music, specifically around engineering, you know, too, as well, not sound engineering, but being a software engineer and stuff as well. 

Adrian Green: Oh, no, we can get into that too. 

Marlon Avery: We can. 

Adrian Green: I mean, my bad, you know, because I was going to start this conversation with my experience with Fruity Loops and. 

Sekou Doumbouya: Being me and, you know. 

Adrian Green: Pre Y2K experience with fruity Loops on my compact computer. 

Sekou Doumbouya: But you definitely paid for that, I'm sure. 

Marlon Avery: Yes. 

Sekou Doumbouya: Pytk. Yeah. 

Adrian Green: I got one word for you, whereas. All right, so. Well, I would say so I just found out about this product or this announcement today and yeah, super impressed of. Out of the other audio tools that are there, I feel like this one, I feel like that. Now. Star wars reference, incoming. But that scene in Phantom Menace where Darth Maul is like pacing back, but kind of, he's him. His fight with Qui Gon and Obi Wan kind of gets broken up. And there's that bit where he's kind of pacing back and forth and waiting for the doors to open. So I said they can resume battle. And that's what I feel like the music industry is doing around this article right now. Like the executives at Universal Music Group and things like that, because what they have done has made music creation. It's given you a scalpel for that. Which I would describe Suno and Udio and the other competitors out there as more of like a battle X. You know, it's more something more blunt that'll get you a song that is cohesive and, you know, together and to a degree, it's not going to be Able to. You can get in there and you can fine tune things as well. But this was fine tuning things with text, which was crazy to me. Like the, in the demo, the way that they iterated through specific sounds where at the end of it, you can have your own unique sound that you created. Like you create that you. You have that sound that's copyright free that I don't even know where to start. It's going to be an interesting time moving forward as far as production goes with just the ability to create tools like this out there. 

Sekou Doumbouya: Yeah. I feel like it's like they're definitely teasing the fact that this is like a research project. It seems right. It just means. It's almost like Nvidia is saying, look what I can do if I would start. Really? Yeah. Oh, just. 

Marlon Avery: Just so everyone knows what a villain story starts. 

Sekou Doumbouya: Exactly. This is. This is literally like a villain arc. I think that like with. I don't know, I was looking at this, this tool. Like there was a. There was this one section where someone was playing the piano and they wanted to add a singing element to the. To the song. And they just said, you know, sing the song. Say a song based off of whatever this context is to the melody that. I just played that right there. That's crazy. It just makes a new. It's like a. It brings a new concept to a one. Like, was it one man show? 

Marlon Avery: Yeah. 

Adrian Green: Whereas before you needed the symbols between your legs, you needed the drum on your back, you needed the whole thing. And. Yeah, the meme. Now it's. 

Marlon Avery: Yeah. You know Ray Charles. 

Adrian Green: Yeah. Oh, you know what, Who I would like to see get a hold of this is Reggie Watts. You know who that is? 

Sekou Doumbouya: Reggie Watts. Oh, yeah, yeah, yeah, I know Reggie Watts is. 

Adrian Green: Yeah, yeah, I would like this. And what was that? He's like a loop musician. So he gets on stage with a loop machine that you can make beats on and create loops and he create whole songs and comedy bits and everything off of it. So it's like kind of like Bo Burnham, but like, you know, kind of really. They're both clever, but, you know, really kind of like a Bo Burnham esque comedian. 

Marlon Avery: Yeah. I want to get. Yeah. So we gotta figure out a way, an audience if you're. If you're. If you know or have a connection. We gotta get like Timberland or something here to like talk about this. What do you see? Because he's already using some of these tools, so. He's already using some of these tools and everything. So I would definitely love to see like his, his, his POB in this, you know, and how has it. I imagine, you know, I'm, I'm no musician or anything, but I imagine the principles of how sped up the work of an engineer or a blog writer, I think it's similar for a musician, you know, who knows how to use it. So it's not simply just kind of replacing, it's kind of like sped up, you know, the, the output really to kind of like, you know, output some of these things. And so, yeah, I think it's, I think if, I think I would say this as a fan, as a builder, I see this as being very interesting. I see this as being very fun. But I also see the power shift behind the scenes, you know, I see the power shift. I see how if you're a music, a music level, you have to be terrified right now, you know, with all these things, stuff happening. Because as we know, like, you know, music labels, they don't typically, you know, invent, you know, they, they play defense. And so this is where you get your Spotify, your Apple Musics, you know, your Pandora's or anything. So for the world. And then typically once, you know, once the music label has typically sued them, they're like, okay, okay, we're done suing now. Let's partner. You know, they did the same thing with TikTok. And so, yeah, I can definitely see how this is shifting. My biggest question, you know, of this or in this is that where does that leave? Where does that leave the artists? Does this give more control to the artists? Now if I'm using, you know, something, if I'm using something like Fugado, does Fugato have to be kind of written into the rights and stuff of the song? You know, stuff now, you know, do they get points of the album, you know, and stuff now? Or is it, is it a one use license? Or I use it and have the license and be able to, you know, claim it as my own as an independent sort of thing. 

Adrian Green: That's a good question. That's a good, that, that's a good set of questions. I was thinking the same thing when we were talking a couple of weeks ago about the UMGs, UMGs partnerships with all these AI groups and that if, which includes devices, includes the people who make beat making devices and things like that. If something like a partnership or, you know, things that you create off of that, if they could, you know, turn on that switch to where. Yeah, because this is made. You're using the AI from These artists then part of, part of the ownership with generating a beat using this, you know, beat pad or whatever that you brought here is going to be, you know, we get a cut of that. 

Marlon Avery: And UMG is nowhere near, not even close. Not even thinking about looking to give any amount of control away way. 

Sekou Doumbouya: Yeah, yeah. 

Marlon Avery: Hey, it's a business. 

Sekou Doumbouya: They're. That's a bit their livelihood. Right. So I think the, the one thing that I think is also just one little take takeaway here. The fact that this is a almost like a research project means that just one of the things that Nvidia is kind of good for, you know, the, the areas where they're strong at, I feel like they invest in heavily. And then these research areas, what they end up doing is not about like creating it, but like we saw, it's. 

Marlon Avery: More about creating, building the infrastructure. 

Sekou Doumbouya: More, more infrastructure. Like hey, we are going to create more ways you can use our product. 

Marlon Avery: Yep. You can build on top of it. 

Sekou Doumbouya: To give you inspiration to me so someone else could create a new vertical. But you know what the is you need our GPUs. This advertisement advertise the AI the scenes. And I think it's, it's super smart. And they publish a research paper on this for anyone to see. They go into detail the different experiments. It's enough information in there. Like if you want to do the same thing, you have a good starting place right now. But guess what? You need some Nvidia GPUs to make it happen. 

Marlon Avery: You know how I view Nvidia? I view Nvidia just like I view the pharmaceutical companies. 

Adrian Green: Or are you okay? So if you're going to comparison, I think you could go into drug dealers. 

Marlon Avery: Watch this. They see a problem, present a solution and here's the solution. Or we partner with other institutions of here. This is a research for the solution and everything. The side effect of the solution, like, like Seku said, you got to use our GPUs. You know, there's always a side effect that rounds them back in that loop. You know, everything. And so yes, you can take this, you know, this solution, everything but the side effects, you know, a lot of times can bring you back, you know, to the source here. And then they kind of keep the cycle, you know, and stuff going. And so yeah, which is, I guess. 

Sekou Doumbouya: In this case, can you say that this is for the good? The fact that they're doing this, they're throwing things back into the ocean for people to pick up and to do things. No GPUs at the end of the day are what you're going to need to use to like do this at any scale. Yeah, you can do other things. They're going to of course appear down the line or there's other like specialized chips you can use for inference and things like that. But like if we want to do it at scale, at large scale, those are the folks that I feel like Nivity really wants. You're going to need to have them. 

Marlon Avery: I think this is a constant flex on Nvidia's part. I think they're at a place right now where they've hidden so many of their goals. They're trying not to get bored. And so they're just, I would argue this probably came out of like an internal hackathon. 

Sekou Doumbouya: Oh sure. I'm sure they have, you know, they have captured so many talented folks over there over the last year with this just their, their rise to being the largest company. They've been gobbling up great engineers in ways that I have not seen in a long time in the Bay Area. And so I am not surprised that they're, they're like kicking out stuff like this. This is the mind share is going over there, I think. 

Marlon Avery: Yeah, this is a constant flex. This is just like, you know, don't say no while to, to piss us off because we'll just, we'll just release this state. Look, just. We create a whole new department, you know. Yeah. This is a constant flex, you know, of who we are, what we've done, what we do. And while we, while we continue to be better than anybody else out there. This is a consistent flex while, while. 

Adrian Green: Chumming the waters, you know what I mean? Chumming the waters. 

Marlon Avery: Yeah. And just like it's. Who, who's who's second on the chip side? Is it Intel? Intel. They have like, I know they have a small market share on the chips. 

Adrian Green: Where's AMD falling at accelerators? 

Sekou Doumbouya: So AMD is up there. Right. But I'm not sure how. It's hard to say. Depends on what you're talking about. If we're talking about inference, then the story is a little bit different. You have the Intel's out there that are still kind of the. Somewhat of the, of the king on that side in some, some fashion. Right. But now there's so many specialized accelerators that different companies are creating. You have. Is it Inferentia? You have Trainium out there on the Amazon side. Azure has their own specialized chips for, for doing both of them. But the thing that Nvidia has that the other folks don't have, since it's a GPU. GPUs and CPUs run on the same machine. So I can get one box that has a CPU on it that can do good, you know, help me with some of my inferential inference. Inference work. Right. And I can use a GPU for training, which means I can run at high capacity within one system and get the most out of it. So it's still like if you're, you know, if you only can, if you only can spend money on like 50 systems, you would probably rather them be able to do both. 

Adrian Green: Yeah. 

Sekou Doumbouya: As opposed to just doing one of those, one of those like feature sets. And that's where I think Nvidia kind of like really, really wins right now. So. 

Adrian Green: Because they can just focus on the one. 

Sekou Doumbouya: Yep. 

Adrian Green: They can focus on the one. Yeah, the one GPU come out with. 

Sekou Doumbouya: A cpu, which here's like this rumbling of them coming up with a cpu. So that will be. That will change the game. Definitely. 

Adrian Green: Interesting. 

Sekou Doumbouya: Yeah. 

Marlon Avery: Okay. Speaking of changing the game. Uber is now leveraging its gig worker platform to enter into AI label business, connecting companies with independent contractors to perform essential tasks for training AI models. This new division called Scale Solution expands Uber Business Model into the rapidly growing field of machine learning. Uber is recruiting workers from various countries including Canada, India, Poland and the US to perform tasks as such as data leveling, testing and localization. And so this move highlights the heating, the hidden human labor behind the AI development. Training AI models often requires numerous workers to compete tedious tasks like labeling images or evaluating chatbot responses. So these tasks are often outsourced to workers in developing countries who are paid low wages. For example, an engineer in India reportedly being paid only $2.37 per sheet for comparing and rating AI generated code solutions. Uber aims to tap into the global workforce by offering platforms for companies to access a diverse pool of workers for their training needs. This isn't Uber's first venture into the world of AI. The company previously invested heavily into self driving cars, but shut down the program after a fatal accident involved one of these vehicles in 2016. It says Uber seeks to capitalize on the booming demand for human intelligence in the development of increasingly sophisticated AI system seu. Uber said they're going to lead this giga gig training AI economy with the power of human in a loop, if you will. What's your thoughts and stuff here? 

Sekou Doumbouya: Yeah, this is an interesting one. It's one of those things where you look at tech, we're all consumers of tech. Out here. And we sometimes we don't think of what it takes to actually get to these wonderful things that we have in front of us. Right. I remember one like, was it years ago we got like, I jumped on the EV train, right. And I was super excited. Oh yeah, I can get this car. You know, it's lower gas, you know, I have to wait pay for. Pay for gas. I'm able to, you know, get access to cool tech and things like that. But then later on you learn that, you know, there's a connection between the lithium ion batteries and the sourcing, which is a large producer of, of cobalt, which is the main, one of the main ingredients for that. It's coming out of the Democratic Republic of Congo. Right. And the fact that US companies, since we're so big or so influential, we can, you know, there's, we can, we can exploit, you know, we can continue to exploit Africa in some ways and you know, these other economies, I feel like we gotta do better. That's. I think that's what it really comes down to. Right. And just because. Well, actually in Uber's case, they have decided that they are going to be the middleman. They are going to be the Uber of. 

Marlon Avery: Pun intended. 

Sekou Doumbouya: Uber of labeling. Yeah. Which for folks who don't know like what labeling essentially is, is when you would go through the, someone would go through the process and say example of one on the image side of this. 

Marlon Avery: We, we've all done it as users already. 

Sekou Doumbouya: Yeah, we've, we've all, we've all done it by saying yes and no on different feeds and things like that. That gives signal back on whether something is good or bad. But in cases where, give you example, like if there, if there's like new imaging technologies that need to be able to detect what a hot dog, what a hot, what a hot dog is, that's what a hot dog is. So solid point has to look at a hot dog, circle the hot dog, put those input values into a system in order to record that this is a hot dog and to build that as the actual training data for that. And because there are always new areas and new things that we want to look at. Like there's always a origin because AI is not creating anything original. It is recreating what has already been that's already there and seeing the patterns that are associated to it. So I know there is a, there was a interview that was looking at on 60 Minutes where they're interviewing some workers over in Kenya who are doing labeling through some third party that other AR providers were using and you know, supposed to getting paid like at least in the, in the, in the report so that the people were reporting that they're getting paid like $2 an hour to go through and do labeling. And it's the type of labeling is like the most challenging version of that, which is the labeling of like, you know, like, what do you call the trust and safety time material? 

Adrian Green: Yeah, to explicit content. 

Sekou Doumbouya: Explicit content. And like, that's that for that, that, you know, it's crazy. It's crazy folks needing to have like therapists and things like that and support. And in countries where those like, availability to services are not quite there, like, I don't know, I feel like you have to remember, it's like all the technology that we're looking at, there's always some human cost is happening right now. And you want to know where your AI comes from. I think that's actually going to be the next thing that's going to come out of this because I think this is just the beginning of this conversation, especially after that 60 Minutes report that came in. Like, you know, the, you have the, you know, USDA labels on our meat. So we know that our meat is organic or grass fed and things like that. We now know that if you buy diamonds, you can, you can specify whether your diamonds are conflict free or not, even though they're still. That's a very murky area. 

Marlon Avery: Yeah. 

Sekou Doumbouya: Yep. Bro. You know, there's at least some attempt, I think we need that over here on the AI side. We need, you want to know that the data set, where it's coming from, not that it's, you know, not just if the data set belonged to someone who has, like, intellectual, you know, ownership of some of that trading data, but also what was the human cost? Yeah, went into that. Right. 

Adrian Green: Went into training. 

Marlon Avery: That. 

Adrian Green: Yeah, that's a good point. 

Marlon Avery: So I'm sorry, go ahead. 

Sekou Doumbouya: No, I think that's the end of my thoughts, but I think that's something that I saw that's like, man, I got such an internal reaction of like, oh, wow, this is terrible. Am I connected to this person? So first it came to my mind, I was like, wait, am I connected to this somehow? Hope not. 

Adrian Green: Yeah. 

Marlon Avery: Yeah. Okay. 

Sekou Doumbouya: Yeah. All right. You got. Oh, it's better out. 

Marlon Avery: Okay. So, yes, I, I do agree that there will, at some point there will be some type of like labeling that, you know, this AI, this AI data has been created safely, you know, or it's been created in, you know, feasible. 

Adrian Green: Human conditions and, or just no Humans were harmed with the creation of this AI. 

Marlon Avery: Yeah. So a couple of things. What's the definition of low wages in a country where the cost of living is already extremely low when you can't afford to. 

Adrian Green: Luke. Cover your bare essentials. 

Marlon Avery: Yeah. 

Adrian Green: Your bare necessity. Yeah. Can't say. 

Marlon Avery: Yeah, but that's. And I'm. And this is, this is no left or right conversation. But I'm just, I'm just like asking general question here. You know, $2. 

Sekou Doumbouya: $2. 

Marlon Avery: $2 or per seat could be equivalent to somebody over here making $20 an hour. You know, anything, you know, in, in, you know, a certain country and everything. And so, you know. Yes, go ahead, go ahead. 

Sekou Doumbouya: It depends, Right? So like a car is still going to cost the amount of a car. It's not like you're going to ship a car over to Kenya and you're going to slash the price because of the economy. 

Marlon Avery: But also talk about, we're talking about, we're talking about places like India too, as well, where there's not a lot of cars. So you got a lot of scooters and motorcycles and things like that. Where is the predominant transportation? 

Sekou Doumbouya: Yeah, but there's global costs and then there's like individual costs. Like, you know, if I'm making, if I'm getting a coconut in India and I live in Kerala where they have tons of coconut trees everywhere. Like. Yeah, it's super cheap. Sure. I'm getting one over here in Nairobi. That's a different story. Right. Because of imports not coming locally from there. So, yes, they. There's local economy. Right. Costs then. But remember, there's also the global layer that's out there. So things are still going to be expensive. Right. Still not going to be able to have the same life. Right. 

Marlon Avery: Right. Okay, so got that. So again, this is no left or right here. This is just genuine question. I do agree, like I said, there will be some type of labeling and stuff there. There too, as well. And yeah, I'm in. I'm an individual, you know. So what was it? What was the big food Netflix documentary that changed people? Like, oh, my God, this is terrible. I'm going vegan. What was the, what was the big one? 

Adrian Green: So many forks over spoons or fork, silver knives. 

Marlon Avery: It was like one large one that like people were like mortified. Everything and stuff of it. 

Sekou Doumbouya: Stop drinking milk and all type of things. Yeah, you're talking about. 

Marlon Avery: Yeah, it so se looking forward. So I know when I hear. But yeah, I, I do think eventually you're probably going to have some Segment of this and everything where, you know, people are kind of like locked into a room and they're just like sitting out putting, you know, identifying pictures, you know, use cases, you know, use case of 1 over 2, things like that. Another thing. And, you know, the work conditions or anything are not going to be favorable. And again, I'm not saying that, you know, this is okay or I agree and things like that. What I am saying is consistently proficiently is a country like the US has also shown, for the most part, they actually don't care. You know, they actually don't care. Like you, you've had, you've had segments of that where definitely it changes people's perspective on, you know, the eating habits, things like that. Do you think? But, you know, where we're also a country, even with those type of videos and documentaries that has been publicized on Netflix, we're still number one in obesity. Like, we still lead in obesity even with like these type of, like, things and stuff like that. So I say I agree on the, There will eventually be some type of labeling. I question how much would people actually care, you know, with that, you know, anything. So I, I questioned that part, you know, stuff of it. The other area of this is, you know, this is, it's, it's hilarious to me. So I used to work at Uber, you know, not as a driver but, you know, as a worker and everything. I used to, I used to work at Uber and everything. And so I'm, I'm kind of terrified to say the woman to say, but yeah, whatever. So I used to work at the Atlanta office and typically it's hilarious to me that Uber is kind of getting to the space and everything because typically at one point when Uber was at its largest scale, it was being shared that the way that Uber platform works is 70% of the wages go to the driver and Uber keeps 30% everything. And that was the, you know, the word of being pushed out and everything. So at the time I was working on like the support, support staff and everything. Marlon Avery was the individual in, at least in the Atlanta segment that figured out that actually that's not how the system works. And I was looking through code I was going through, I was like figuring out. And so I was like, oh, this is not how this is work. And I'm not sure if they changed it or how it works. How it actually works is Uber evaluates Chris algorithm based off your distance, your time all algorithm and it gives a, a guess of how much the ride is going to be. You think for the rider, it gives a guess. And so say for example, you know, say for example, what the health. Yes, thank you. That's what that was the thing called. So it gives a guess. So say for example, it's a hundred dollars, you know, it's a hundred dollars. Right. So it gives a guess to the rider. The ride is going to cost $100. And on the driver side or anything, it does the same like algorithm and it says we consume. We, we guess that basically this ride is going to cost $60. Everything is going to cost, it's going to bust you, you make $60 and stuff on that. Really? So if you got $60 over here, $100 over here, where's that difference? And typically it's not that large, but I'm giving a good example. Right, so the rider still pays $100 and then the driver gets 70% of the 60. So Uber still takes 30% of the 60 and they keep that middle. 

Sekou Doumbouya: Ah. 

Marlon Avery: Yeah. We're the longest near thing, you know. So I'm sitting out to say is that it's, it's hilarious to me that they're starting to kind of get into this. Let me connect this individual to this individual, everything and I'm gonna take middle of it. You think it's the same? You know, yeah, it's kind of saying the business model and everything. So there could be a lot of issues and everything with it. So I, I don't, I don't. This is interesting, you know, if I, I hate to admit it's kind of genius on Uber's part, you know, and it's, it's a, it's a very, it's a very, very genius move. There definitely going to be some legal issues and stuff around it and everything to find human beings to do training, you know, for set, you know, data sets and everything. But as users, we've all logged into Google and Google says, hey, select the squares where the bicycles are, you know, and everything. And so that is training. You know, you're training, you're training an image model. It's up in that and everything. So that's training. And so now they'd be looking for people to do this full time and everything. So whatever it be. Third thing is, I find it interesting that Uber and they've done pretty well in this area, I find it interesting that they're predicting that we will need more data sets when there are certain areas that say, like, hey, we don't actually need, we got all data sets, or you can use synthetic data, you know, and Uber's saying like ah, but that's not true and everything we're going to need more data section and particularly for your specific use case areas and things like that. And so I find this very interesting which also means that the foundation of what we thought we were at the bottom of the first until like this build of gen AI and then hey, we're ready to take off and everything. It looks like if Uber's right, it looks like we're probably in the middle of the first. At the bottom of the first. Because if Uber's predicting that you're actually going to need more data set that you actually think you're going to need to get very much more outcomes, everything, which also means you need more inference, more GPUs, you know, all different things and stuff as well. And so now you're going to need more cloud computing, you need all more of these things which you think going to use anything. And so now they found a network if you will, in the middle and saying like, hey, this is, this is one of the use case, one of the saws that's going to help us kind of get from the bottom of the first to the top of the second or where it may be. I find this as annoyingly brilliant on their part. I find it very, very interesting and stuff. And so I'm definitely, I'm definitely follow this more. Yeah, cool. 

Adrian Green: Cool. Yeah, I think that it's interesting because just from a PR standpoint like Ubers, when we're talking about the 60 Minutes piece and we're talking about the conversation with around Uber actually maybe sourcing the people who will be, you know, working for these most likely super low rate wages if they're outside of the US and so it kind of like it looks strange in the PR the PR ness public relations ness of it because you with me, you know, when I think of Uber, I think of their scandals just as well as I think of how convenient that it is. You know, because it's just there's been so many, I've had people that are. 

Marlon Avery: You know, like, like McDonald's. 

Adrian Green: Yeah, it's, you know, but more so with Uber, I mean the McDonald's right now that that comes to mind, I'll be honest with you. Yeah, so, but you know, it's, it's tough because if in order for this to be profitable, you have to be putting human beings through this kind of thing and from a standpoint of just, I don't know, Wait, human beings between. 

Marlon Avery: What type of thing? Elaborate. 

Adrian Green: Well, there's Gotta be human beings in this that are involved in this. So the human beings are going to be there that are going to be looking at the images, labeling the things and the things that are in the image. In order to train the AI model in doing that, they're also, you know, training it on explicit images and things like, you know, dismemberment and all kinds of crazy stuff going on where that they have to look at and identify as such. They're the human cost that goes into that. People think about that. Now, going back to the cell phone conversation and the battery conversation, I think that, you know, we. It was not rare to have that conversation normally amongst my friends. Just like, oh, man, you guys think about like the global impact about what is in your phone, you know, and we, we're all kind of like humbled by it a little bit. Like, yeah, kind of. It's really bad. Like at the end of this phone is someone that's having to go through basically, in some cases a warlord or some bad, you know, situations where they're. It's close to slavery, you know, the way that they have to operate. So the PR situation where Uber is inserting itself and being the kind of intermediary contractor, just like the, in the 60 Minutes article, they named the company that the people work for. That company is paid by, you know, whatever AI company, Google or whatever. But that person in this country works for them. Okay? And this person is saying, you know, the company is saying, we need. You're not working fast enough. We need to double your output and we need to give you. We need you to double your output or increase your output in less time, constantly while getting a low wage, while being traumatized. It's rough. It's a rough kind of cost to it. And I agree with you, Marlon too. When you think about anything outside of the U.S. it's hard to the average U.S. citizen, they're really not going to really. It's not going to be their top of mind when they go to the store about the. Whatever human cost that is with that. But I think for those, like in the know, for those that are working, you know, working in the, in, in the industry, I mean, it's definitely. Which isn't the large portion of the US population. They would. They probably have no idea. You know, I don't know. It does, it does kind of kind of make you think. And it makes me question, really, Ubers, why is it Because Uber already has inroads into these, like, kind of, you know, India, Poland, Nicaragua, places like that where you know, they can just use their existing infrastructure there to stand up a contractor business or something like that. Is that it? 

Sekou Doumbouya: So that, that actually, so I'll ask you, you touched on something really interesting here. So I always gotta remember that Uber, when you look at their businesses, they're a global business, they're huge. When, if you travel outside the country and you're getting a cab or something like that, Uber is part of that process. So yes, they have, they have connections within, they have like offices within India. They have infrastructure already set up, set up for that. I would actually, I would actually say the fact that Uber is, wants to be the actual company in the middle and it is the American country, I feel like that actually may increase the standards as opposed to reduce them just because, you know, there's different, it's a different level of liability if like American company is going through and directly exploiting bingo countries as opposed to like some, you know, shell entity that operating in these countries that can just exploit and take advantage of that fact that there is a super high unemployment within these. 

Adrian Green: Countries to 60% in Kenya they were. 

Sekou Doumbouya: Saying, yeah, 60% UN unemployed employment. So and essentially if people complain, these shell companies are pretty much, you know, pick. They would just shut down their offices and go somewhere else rather than fixing the problem. 

Adrian Green: It's totally right. 

Sekou Doumbouya: It's almost like, yeah, it's almost like you're being held hostage and you're going to be retired against in that case. So that's actually, that is actually the huge plus and their business because they can't. 

Adrian Green: Uber couldn't just shut down and then reopen again like any of these other companies could. 

Sekou Doumbouya: No, they have an international brand. 

Adrian Green: Yeah. 

Sekou Doumbouya: They do this like they operate in India. You don't hear the same stories from, from like from things that they do over over there at least. I have never heard any stories that sound like that. So yeah, actually on think about like I think the overall training aspect that was that we were seeing at least in the 60 minute side like that definitely horrified me. But the fact that Uber wants to be the company in the middle makes me like I can at least have a place to put my hope. Right. Because it still has to be done. But I would rather it has some type of like, you know, accountability. Accountability. And that means, you know, they're probably not going to just do it in India, they're going to be doing it over here too. Right? They're going to be doing it at scale in some ways. So yeah, they, they're doing a good they're good at doing that at a global market. 

Marlon Avery: So which is pertains me to my original question. What is the definition of low wages to a place that already has a, you know, low living arrangement? You know, it's, it's that the, the way Uber keeps themselves mechanical is a strong word for them. The way Uber keeps themselves clean is their initial language of how they already worded this, which is AI labor and building businesses, connecting companies to, with independent contractors, which is giving them opportunity, giving them a choice, giving them an option. Everything you are, you are an independent. You're IC in this area and you're coming to our service, our platform sign as an ic everything and then you're automatically agreeing to whatever percentage base, you know, allocation that will be distributed based off of your efforts and your work and everything. It's the same model. It's the same model as the vehicle. It's simply now you're just using human beings to get to take a data set from point A to point B. 

Adrian Green: And that's what sucks too is because they're contractors, you're not setting them up to actually have a career, you're setting them up as a contractor. 

Marlon Avery: And that's the whole point. It's the, what's this? The gig economy. That's the whole point. It's shifting away from that. It's the whole point. It's the gig economy. 

Adrian Green: If you were to, if you have a country that's closer to third world and then you're going to roll out the gig or something, that's going to contribute to the gig economy. I really think that that's going to do nothing but just keep the economy low, keep the, keep the performance low because there's no upward mobility in that. 

Sekou Doumbouya: Cause it's not like people are like if the place has, if a country has 60% unemployment rate, having folks having contracting positions where they're moving in and out of, you know, joblessness is, that doesn't improve the prove the situation. Also your ability to say no in that situation is almost like, doesn't exist. It's impossible. So you have to feed a family. Right? That means that if there is a situation where something is, something not good is happening, like your ability to speak up in that case is like non existent. It's like it's the lot of the set of the type of leverage that side inside there I think is that's what, that's what makes it a little bit more exploitive. And when you're not actually paying what is actually a human Living wage. 

Adrian Green: Yeah. 

Sekou Doumbouya: In the country or setting up, setting up folks to they just take advantage of cheap labor. Really? Yes. You know, and that's, that's a. Hey tell you that's how, that's how you know a lot of folks came to United States. 

Marlon Avery: That's how this country is built. For sure. 

Sekou Doumbouya: She Cheap labor, right? 

Adrian Green: I mean I think it, it was below cheap, but yeah. Agreed, agreed. 

Marlon Avery: Yes. I okay, so I, I'm a disagree with the initial point you just made which is this keeps the, the wage is low. I'm a disagree in some of the area. Independent contractor. Let's, let's take the vehicle standpoint of where Uber stands today within America. The. Some individuals found a way to make it a career. Some individuals, like when I was working at Uber we had some drivers that was making six figures and even to us we're like bro, how are you doing this? You know, and everything. And so it was a small percent, you know, for sure, everything. But we had some drivers who figured out we had some drivers and everything who were able to, you know, use this as a side income so they can go buy their house and then they stop driving. You know, we had some drivers that were just using in between roles, you know, they, they land their own, you know, everything. What it does is if you have an area where it is 40% unemployment, it gives opportunity with the alternative still being nothing and everything. And so yes, it is a form of exploiting but I think it also too comes back to the individual on what they do with that because some people can use this as a stepping stone and the ground like man, okay, I'm at the bottom, but if I can just get a step free, I can figure it out. You know, step four, five, six and stuff that may be. And then some individuals will be like, hey, you know guys. And I'm gonna wait until they decide to pay me more. I'm gonna wait until they decide to do a like that is kind of what we do and who we are even in this country and stuff. You're gonna have people who took their job from McDonald's and everything and used that as a funding to start their long cut lawn care business. Everything. It's the opportunity, everything. So I, while, while it's. I hate to I wow. This is really hard to say. While I understand the wow, okay. While I understand what we're seeing on opportunity to a place of that has 40% unemployment, I also would say it's also not their sole responsibility to come in and just solve the whole thing. I think you come in and you give opportunity as a. As a stepping ground. Now. Yes. They figure out a way how to leverage that to get what they need, you know, to as well. Or get me an opportunity to build a new business model, a new division as well. But I think in this area, you will have people, which is. Wow, that was. That we used at $2.37 per set as a stepping ground and be able to put their child in that middle school that all they've been trying to get to. They just couldn't afford to do so is a stepping ground. I don't think it's their sole responsibility to come in and just solve this globally. In some of these countries, I see this as. Here's the opportunity as a ic. Use it or not. 

Adrian Green: Yeah. I mean, I think that's definitely one. 

Marlon Avery: That was tough to say. 

Adrian Green: Yeah. Harsh perspective. 

Marlon Avery: Yeah. 

Adrian Green: Because it's basically like you either take it. It's better than nothing. You know, at the end of the day. 

Sekou Doumbouya: Yeah. 

Adrian Green: Even at McDonald's, you can move up. You can't move up when you're labeling the images like it, like it's like it's going to be very. It's not really a lot of. When it comes to contractor work, you're not. You wouldn't be a contractor for McDonald's. You'd be employee. So there would be certain benefits and things that kind of go along with it. And it does kind of seem like, I mean, with a company that has that much money and it could, you know, one thing that could be transformed. Can transformative number one is like, say, I don't know if this will be possible there, but just like benefits or something like that, like give them something to which they can, you know, maybe help that. 

Marlon Avery: That took years on the driver's side. Years. It took years. Yeah. 

Adrian Green: If they. If they are able to do that. And really. Because if you do that, you can change their lives. I mean, no matter the wage. If it was benefits involved. 

Marlon Avery: Yeah. 

Adrian Green: So it's. And they're able to do that. I mean, it could be. It could. I don't know if they're able to do it, but it could be true that they're able to do that. And you know, but if it does, to me, it kind of kneecaps the possibilities of you. It just further just like kind of makes a truth resonate that there's not really a lot of opportunities there to work, you know. 

Sekou Doumbouya: Yeah. 

Marlon Avery: If I'm not mistaken too, I believe also the healthcare, the benefits part for the driver's side. I believe that was forced by the government on for Uber. I believe they lost like a court case or something like that. And it was, it was like they, they were kind of getting out of. Because I'm saying they, that basically the drivers wasn't recognized as employees. So they did, they didn't have a need to do so. And then yeah, the court said yeah, whatever, you gotta figure this out. Yeah, okay. 

Adrian Green: And that's the ul. Ain't no court too. Ain't no courts over there. It's gonna be like, you know. 

Marlon Avery: Yeah. 

Sekou Doumbouya: Up. Yeah. 

Marlon Avery: Okay cool. So you had any last top thoughts on that one? 

Sekou Doumbouya: No, no extra thoughts. It's actually curious about this, this next topic that's about to come up here. 

Marlon Avery: Well we, we got, we got, we got a few and stuff here. I'm. I want to, I'm gonna save this, I'm save one of them for the last stuff here. Lately there has been a surge of downloads, a surge of, you know, people switching to this new platform called Blue sky, which is now the new competitor of Twitter and Threads. I'm beyond you guys. It took me forever to figure out their Blue sky that was like actually competitor Twitter. Anyway, so Blue sky now has some concerns around their third party party. Third party AI training it says Despite Blue sky stands against this training its own AI system on users data. A recent incident highlighted the vulnerability of the public post to a third party scraping and used for AI development. A machine learning. A machine learning librarian at hugging face downloaded 1 million public blue sky posts using the platforms platform's API for machine learning research and published the data set. While the data was later removed due to uncertain controversies, the incident underscores the potential for the public BlueSky's data to be used for purpose beyond the platform's control. BlueSky acknowledges its limitations in enforcing user consent preferences outside of its own system and highlights the responsibility of third party developers to respect these preferences. It's funny, the company is exploring ways to enable users to communicate their consent preferences externally, but the effectiveness of such measures remain unclear. This situation raises crucial concerns about the day where privacy controls in the age of social media and ad development. While Blue sky is gaining popularity, it now faces some same scrutiny as major platforms regarding data usage and ethical implications of AI training. Sekou, we'll start with you. Blue sky already underneath the fire seat. 

Sekou Doumbouya: Oh man, I was a little sad about this one. So just a little background like was it the other night I was talking with my wife about this about Blue sky, showing her how the, how the feeds work inside of it. You can like have a feed or even create your little own feed that like looks the way you want to look based off of regexes of, you know, posts that you would want to capture and adding it into bluesky. And I got her signed up, got her using it. We're about to get rid of our, you know, X accounts. Right. And then today I was going through and I saw this, I was like, oh man, this feels like, you know, I feel like they're doing a good deed by like creating this company and, you know, setting them, setting it up so that like, what do they call? They said they're billionaire proof. That was one of the things the CEO said. Yeah, because like, if you wanted to like buy out Blue sky, all of their things that actually make them unique are all open source and available out there. So you can just create another one if you want to and reuse some of the, you know, and, and you don't. You just know where to like get full ownership of it. Right. So I was super excited about this. When I saw this, I was like, oh man, this feels like, what are they, what's the saying? No, no good deed goes unpunished before. That's what it feels like. Right? This, this act of kindness that they have put together is complete, completely, is kind of like backfired in a way because one of the things that happened on X when it was taken over is they got rid of your ability to create just integrations to connect to a stream, to build your own web ui. If you want your web UI to look different, you wanted to set up your fees in a particular way, you can build something for that so they have a public API that you can just use. And it's, you know, it's awesome when a company provides that because you can like, you know, I feel like I got a lot of like my experience, you know, doing integration. The fact that someone had a public API that I can actually build things around. Right, yeah. So it's super sad and the fact that this person has turned it off great. But like, not everyone is going to be hitting up this, hitting up their stream to do exactly what you just turned off. And now that that has been published everywhere, like what are they going to do? They have to put some, I don't know what, put some authentication around that, like make it so you have to have a signup page in order to get an API token and prove that you're not going to use it for training. Like, I don't what can you do. It's not a lot of good options inside of here. Yeah. So I don't know. That's, that's how I'm feeling right now. It's a little, little down. But I'm still a Blue sky user. I don't think I'm going back to X. And I never really understood threads. I don't know, might be the, you know, my generation that I'm in. But what, what, what is happening here? You know, it's truth. I never truly understood Instagram, to tell you truth. I'm just being up front. Yeah, I have no idea. If I cannot search it, why does it exist? 

Adrian Green: All right, yeah, you got their ar. Your ARP membership is going to be in the mail. No, you already got one. 

Sekou Doumbouya: I didn't have one. 

Adrian Green: All right, well then you're, you're, you're duping a new set of golf clubs. 

Marlon Avery: Them when you re up. 

Sekou Doumbouya: By the way, the Black sky over 40 is where it's at. I'm just saying. Okay, that's, that's the feed to jump on. 

Adrian Green: Okay. Black sky over 40. 

Sekou Doumbouya: Yeah. 

Adrian Green: Okay, cool, cool, cool, cool. Well, yeah, yeah, yeah, we'll circle, we'll circle back on that later. Yeah, so, so I want to give you like, what I have to say is like, I want to just like, you know, give you like another perspective because Twitter, all the way up until the Takeover API was wide open. That meant that you can do the exact same thing that you can do. It was, it was wide open. That data firehose API, they have the exact same one for Twitter where you can get everything that was open. And because, and I know that because I use that, you know, I was building applications at the time like Geo Loc, kind of like Geographic Aware applications to pull social media posts from different, you know, geographic radiuses. So I used it then and it was awesome. And what I have to say is that I like the fact that it's open like this because I would like to. There are times where I want to build off of the data that's on X now. So I would like to set up, you know, scraping operations or something like that. What will I have to scrape now because of what they forced me to do? Okay, be sure to digitize my face too and, you know, learn my voice too. But yeah, I, it, it enabled me to really plug into Twitter and, you know, really train my own model. Like train, train my own stuff. Like do. Do my own training myself. Which if you are posting on social media, your data's out There anyway, like, it's scrapeable. Like, so that's Instagram, Facebook, it's everywhere. Like you can scrape all of that stuff in an automated fashion to backfill whatever Lex Luthor vector database you got on the inside the volcano. You know what I mean? You can do that already. So this seems to me like a little bit of much to do about nothing. Kind of like, okay, you're kind of blowing up something that's like, is a standard and according to me should be. Should be kind of normalized. The kind of the ability for us to pull consensus, you know, sentiment data ourselves without. Because, you know, X is doing it. X is all using that to feed their own system just internally. Okay. So I like to have access to that too. So I say I'm all for it, keep it open. And right now, the way that Blue sky was launched was really probably their uptick in membership recently. It probably has to do around the election where it seems that people made a kind of like a blue exit, if you will like to the Blue sky platform. After the election there was a lot of stuff like that. But now me personally, I'm thinking like, okay, well now this is just another polarized area. So I'm still on X. I'm still an X user, you know, still checking articles and using that as part of my like daily news feed. Because I almost feel like I don't want to jump to, you know, not to get into politics here. But I'm not trying to jump to like the liberal parlor at any point. You know what I mean? I like to stay in the middle with my, you know, news. And my feet on X has not become radicalized. I'm sorry to hear that about a lot of people, but my feet on X not so I. It is there which I will remain. However, if there will be a mass exodus, I will gladly join Blue sky with my fellow brethren and enjoy the hopefully, hopefully still existing firehose API. 

Sekou Doumbouya: Yeah, yeah. Anna tell you for my experience on not. Not trying to convert you, by the way, I'm not trying to convert you, but just saying the fact that you can choose your own algorithm if you want to have algorithm that looks like cats. Like you are a cat person. You love seeing pictures of cats. That makes your day. You can add a feed which is, you know, maps to an algorithm that is just for that. And every day you just go and you pick up your Blue sky and just look at your cat feed. New cats will be there every day for you. It goes for anything that you're Interested? I have one for. Was it for AWS something? It's like AWS enthusiasts or something like that. And I can go through and just look at that and just see what the new things that are happening with that particular subject is almost like. I don't know, it's like a mixture of like, almost like RSS speeds. Right. There were RSS feeds from back in the day. I have, like, RSS reader from all your different sources. You get all your cool news that you want to see. Like, I'm doing that essentially at, At a social media level now. So there's a. What they call Black sky, which is. It's like, you know, Black Twitter, which existed, but, like, you had to, like, be plugged into the different food folks in order to. 

Adrian Green: Yeah. And, you know. 

Sekou Doumbouya: Yeah, they weren't. They were not, like, for me. 

Adrian Green: And, you know, I found out about Juneteenth in my 30s. All right. And then, you know, so I have a history of being very late to, to whatever the black, the black people are doing, like, where they're at. When they told me about Black Twitter, I thought it was something I could just hop on. And then I figured out it was a series of accounts that I should have already been following. 

Sekou Doumbouya: Yeah. 

Adrian Green: You know, serious. 

Sekou Doumbouya: But now it's a, it is a feed which is hilarious. Like, I just go in there and we'll laugh, like, and it's etho. It's. It connects to the culture. Well, if I want to laugh to something that's connected to the culture or, you know, see what the, what the discord is, I can go to the Black sky feed if I want. If I want to go back to my cat pictures, I can do that. I can just flip back and forth. But that choice right now, I think folks are realizing that this exists now. If you want. It does mean, like, if you want to be an echo chamber, you can be an echo chamber. Like, hardcore. Yeah. But I don't know. 

Adrian Green: The bubbles are there. 

Sekou Doumbouya: I'm having a good time. I got my wife to convert. That's my first real conversation convert so far. 

Adrian Green: So I'm drinking the Kool Aid over here too. 

Sekou Doumbouya: How about you, Marlon? Are you using it? Are you using Blue sky yet? Have you, have you. 

Marlon Avery: No, I'm, I'm not using. I've heard about it several times. I, I, at this current moment, I cannot add another thing to my device that, that will distract me, so. Oh, at this current time right now, I'm refusing to do so and stuff. And so I'm looking for this maybe it right here. Okay. So I don't know. I feel like this is. This is in some form, you know, with. With Blue sky and all different things like that and, you know, third party AI training and things like that. I feel like this, all this is kind of like repeating history. So going back to what Adrian said too, as well, at one point, Twitter did close API long before, before Elon acquired it yourself and everything. And so I want to say it was like 2015. It was like 2015 or something like that when Facebook roundt. Facebook closed theirs. And so if you remember, it was. 

Adrian Green: Still open because I was still working on it. 

Marlon Avery: No, no. So. Well, no, no, no, no. They closed segments of it and everything. Because remember when Twitter started opening, it had an open RSS feed, so they had an open RSS feed and everything. And they closed that and they started to. They started, you know, they started to close segments and stuff of it and everything. And developers was getting pissed because people was building companies and businesses off of the open access of it. 

Adrian Green: Facebook shut down. 

Marlon Avery: But Chris Dixon wrote about this in his book Rewrite Own and everything. This is kind of as I was looking for like the deep dive and stuff I'm into as well, because basically both platforms started to kind of like start to mirror. Mirror stuff, mirror itself and everything in certain different ways and everything. Facebook was like closing large portions of it, but it was, it was more noticeable because Facebook is such a, like, large platform. Twitter was doing. Starting to do the same thing and just smaller segments, everything and stuff of it. So they started with the RSS feed and they started, they started coming closely, you know, all these other different platforms as it kind of move along. 

Adrian Green: So. 

Marlon Avery: It feels like history started to repeat itself and it started to repeat itself and everything. It's this world and everything. And it's like you can say all day, it seems like every time somebody says they want to do something right for humanity, you know, everything, it is a second thought after they figure out a way how to lace their pockets on in, you know, in Mr. Way everything. And so with this everything, with this blue sky and everything example that we're kind of talking, you know, talking about as well, you know, I, I just. It's not surprising. I know Seiko said this hurt his feelings everything, but it's just not surprising and everything, you know, in this type of area, just because it's just, this is the way of the land. You know, right now somebody's trying to. Somebody got access to 1 million public, you know, blue sky poster thing. Because now they're trying to figure out how to train, you know, their models for the, their use case for their thing. You know, they're trying to build on top of this to get a certain results or you know, fine tune or support maybe, you know, stuff in this area. And it's just like with Blue sky kind of already been here, done that type of thing. It's like, bro, don't tell me that you didn't see this coming, you know. Oh, what they did, Rex, Wow, that's a new thing. Like, come on guy, don't, don't tell me that you didn't see this coming, you know, don't tell me that. Hey, watch this. Don't tell me and don't even try to convince me that you're not already doing this yourself. 

Adrian Green: Yeah, already happening. 

Sekou Doumbouya: Yeah, yeah, they did, they have their commitment that they will not do training of posts on. They want to do AI training. They have, they had this commitment that they actually published. It was a couple, was it a couple weeks ago around that they say actually how long ago was that? 

Adrian Green: Around November 15th. 

Sekou Doumbouya: 15Th. Wow, that's, that's very timely that this got reported. So I think you're thinking, right, like they, they knew that what was going to happen. This is all publicly available information. I think the ch. The challenge is that makes training on this information so much different is the fact that, you know, one day maybe someone will, they'll have enough data to about how I speak on Blue Sky. They say, hey, do X and do write this article in the style of suku. Right? That's, you know, not that, not that just for me, that's something I think about. But you know what, I still have to remember that back in 2000, was it 2001, the idea that you're posting to a public forum was taboo. It's like, oh, you gotta post it. Oh, someone can steal your, steal your message about, you know, the cake that you want to bake. Right? Yeah, like that was a, that was where we were. We didn't want folks to know your username. And it went, oh man, don't just share my username. That's. They can do all type of terrible things. They have my username. That was 2001. Look where we're at now. Yeah, the world's different. These things are going to shift. So I think what can and cannot be trained in AI needs to kind of go through its own moment. But still, I feel uncomfortable still. So yeah, that's why I like, you know, I will. It's Almost like when it comes to ads, like if you give me an option to pay money to not see ads, I will pay the money. If you give me an option to not have my data trained, I will pay money not to have my data trained. Like that's just, that's just how I operate. As long as I have the ability to do that, that's what I do. 

Marlon Avery: Yeah, John, John there in the comments said, John said that he has 11k followers on Instagram and he still don't understand it. He said they're all changing the algorithm. 

Adrian Green: Yeah, it's true, John. 

Marlon Avery: I'm. I'm going to take that and use you, my friend. Speaking of changing, Zoom has said now they are making a change. Zoom said we are now a AI first work platform and not a video platform. It says Zoom Communication has officially is dropping the video from its name, rebranding as simply Zoom Communications to reflect its evolution into a broader AI work platform for a human connection. So this change signals a share from its focus on video conference which propel its growth during the pandemic. A pure comprehensive suit suite of tools for hybrid work environments. While Zoom experienced explosive growth in 2020, its revenue projections have since declined as worker returns to the offices and face competition for larger companies like Google, Microsoft, Slack and so on. To counter this trend, Zoom has expanded its offers beyond video conferencing, launching Zoom team chat and a full suite Zoom workforce solutions that includes productivity apps and business email, client and more. The CEO Eric Yon emphasizes that Zoom commitment to leveraging AI to automate the work process as approved efficiently. He highlights Zoom's AI Companion 2.0 launched in October with features like improving summarizations and assistant tools. Yan believes that AI can ultimately lead to a four day work week by automating, tying, consuming interesting tasks and freeing up employees to focus on more meaningful work. The rebranding and focus of AI represents Zoom efforts to adapt the evolving landscape of work and remaining a leader in communication technology space. So I'm gonna start off with this one guys. I see this as I see this as a smart play from Zoom part. I see Zoom kind of trying to enter into the atmosphere of becoming a competitor to Google versus just being competitor to Google Meets. I see this now they're starting to say like hey, we dominate the video market. And so now we start to expand our wings especially now has people who are shifted back to work, you know, more in office, you know, more interactions, things like that. So I said it's a smart move, you know. So the Zoom part It looks like they even mentioned just say, hey, we want to start offering business email, you know, for client productivity apps, you know, as well. And so, yeah, I mean, I also see not only Zoom doing this, but also, if you guys remember one of our first episode we talked about Slack, you know, are doing to kind of get into sex, they're going to start to, you know, introduce more AI productivity apps. Same thing with Salesforce, you know, too, as well. And so, yes, I see this now as a kind of business suite of opportunities that we're. Zoom is starting to see. And I think it's. I think it was a smart move, I think. And so we'll see, you know, now who kind of like, you know, figure out who dominates in marketing, you know, who can be, you know, the most eccentric platform, you know, for businesses, small. Small businesses, large enterprises, you know, and Zoom feels like, you know, they have. They already have your video, you know, stuff here. You know, why not use this for email? Why not use this for, you know, our C suite, you know, C suite tools and everything. Why not use this for, you know, productivity tools and everything? And so I think it'd be interesting. I think it'd be interesting to see what the consumer thinks, you know, would the consumer take Zoom seriously as we're, you know, Zoom says like, hey, by the way, we've developed this. We developed this, you know, this email client, you know, we develop, you know, this tool to help you write grants faster, you know, you know, with the consumer, you know, take Zoom seriously, you know, So I guess we'll. Guess we'll see here. Yep. 

Adrian Green: I. Yeah, I think that. I think it's cool, but I also think that this is like, after 2020, Zoom, the C suite, people at Zoom basically have been on like a vacation, and they come back tanned, and they come back like, okay, let's. Four years later, four years later, slipped at the wall. And let's figure out what are the things that we said that maybe possibly we're gonna get a flight back. That's what being competitive pre 2020. Let's look at that board and let's do one of those. And then they took a dart and they said, oh, yeah, Slack, Boom. Team. Team chat. Boom. And yeah, with the AI stuff. Yeah, let's put in an assistant that'll translate. Meanwhile, there's so many plugins to Zoom that do exactly this. 

Marlon Avery: So they're great point. 

Adrian Green: They're like, working a little bit backwards where it's like, okay, well, working against their partners. Working against their partners, really, because what I would see Zoom is when you have a product that you have integrations for, you should. You could invest into those integrations, or you could look at the existing integrations and be like, yeah, we'll just do that. And I would have probably liked to see, probably Zoom go more in that direction of just really developing its integrations further in that area, rather than the shift to AI. Because where they're going with it and the avenues that they're going with it, like, say, like with the team chat that is going to have. That's going to be a product that's going to be hard to differentiate itself from Slack as far as capabilities out of the box, unless I'm missing something. And when it comes to the Zoom AI companion we have, I mean, I am under the impression that at any point there is a team of. I mean, there is. There's a mass of, you know, your Guilfoyles in Silicon Valley, your people that are just always. That are just, you know, working on AI companions themselves. So it seems a little bit unfocused and uninspired to go to be like, oh, we're an AI company because we have things that do AI stuff now where what really is going to happen, I think is going to be a merger between them and Slack or them and one of the other competitors or one of the other platforms like that that are kind of in this space where they have reached the plateau of user acquisition, you know, kind of. Or they plateaued there. So they're going to have to figure out either ways to innovate or who to buy in order to get more market share. 

Marlon Avery: The merch, the merger thing is really interesting because literally I was just. They're just Googling Zoom's value. They have. They have a $27 billion market cap. And then I also just looked at Slack as well. They have like a 25 billion, you know, so I think the merger thing, because I was looking basically, who can Zoom acquire, You know, who could acquire Zoom and everything, instead of, like, trying to build out this width of, you know, tools and, you know, everything and trying to like, you know. So I think that. I think the merger thing is interesting. 

Adrian Green: I'm not sure. 

Marlon Avery: Yeah, that's ill. Now, which. 

Sekou Doumbouya: Well, you know what? Member Slack got acquired by Salesforce, so it's a border. I think you might be looking at their frozen market cap, because they don't. They're. There's. There are now. Salesforce now. So. So I should actually, you know, just from. Just from this experience, you know, seeing how These giant enterprise companies work like they're, they're like, how am I gonna explain it? If you make a competing product on any of the verticals and you're to a, to an enterprise company like a Salesforce or Slack, the immediate, the first thing that happened is like everyone will uninstall soon across the board because now you're competing. They don't want to give you any money to fuel you, which I feel like they're going to lose customers very quickly through that. But Zoom, I feel like they really work at any levels. Like all the church programs are using Zoom. Like all of the family meetups that folks are having, they're not using Hangouts, they're using Zoom. Zoom is that, that is the Go to the Go to app. The fact that they're as different like workspace, Workplace things is interesting. I don't think they're going to get like more usage out of those customers. What they're trying to do, they're, they're trying to directly compete with Slack and they're trying to directly compete with Office 365 right now. Because Office really I think that is the target because Office 365 is eating a lot of folks lunch because everything is bundled in. Even if you don't want Team G teams. 

Marlon Avery: Yeah. Now you got your co pilot. 

Sekou Doumbouya: Get the co pilot that's in there. But I don't know. I like the Zoom product. I like actually notice I like their ui. I like the UI on Zoom. I like how they arrange their buttons. Everything is a little bit more. Feels like a little bit more polished. I was looking at the Workplace, the Zoom Workplace thing that they're putting together and the docs situation. It looks cool. I say it looks cool. I might give it a try. Maybe this like they're competing against Notion now. Like I'm sure Notion was using. If Notion was using Zoom they might have to think about should they use Zoom. Do you want to feed your competitor who's going to try to destroy you? 

Adrian Green: Right, yeah. 

Marlon Avery: So these are all calculations that can be acquisition. 

Sekou Doumbouya: But it looks good. I got to say it looks good. So I'm going to try it and see how see it kind of works and things like that. My, my only request is to make sure that if you're going to have integrated, integrated system that has all these components in there, like make it easy to search documents. That's the only thing I ask. Everyone fails at this having to get like third party things in order to help you search for documents that you Create. Yeah, like please don't, don't mess that up. If you can get that right, you might actually get an AC user. So. But what else is in there? Oh yeah, so the AI companion. That's also pretty interesting. They. 

Marlon Avery: They're. 

Sekou Doumbouya: I think they're doing. They're a little bit ahead of time with some other folks on being able to like, you know, take a conversation as you're recording and it should do something useful for it and give you feedback and meeting notes and stuff like that. I know there's like Otter EI that's out there that's kind of like very good Nucleus on this space. But Zoom is on. Zoom site is really good. It's very well integrated. 

Marlon Avery: But also too. One of the, one of the, one of the controversies with Zoom is they were going to use your data to train their models if you like the. Or not there was that. 

Sekou Doumbouya: That was. That's so that's, that's. Hopefully they, they've backtrack on that. 

Marlon Avery: They will send back thing because they. 

Sekou Doumbouya: Will not get enterprise companies for that. They are very. That's something, you know, folks want to protect Their, their IP and their, their knowledge base that is Documents is really, I think it's an important, you know, that's it. It's really important for a, for them. So I don't, I don't see allowing having that forced upon users is not a great way to get to. 

Marlon Avery: Oh, so like they rescinded it. It says Zoom says it. It will not train AI on calls without your consent, but other data of yours is fair game. 

Sekou Doumbouya: Oh no. Oh no. Okay. Oh, and hopefully get that, that straighten out that might stop me from using, you know. But we'll see. At least from what I'm seeing the ability to like integrate. It looks like you can like write on documents at the same time. This shows you exactly where the person is within the document and you both have a chat window on the side. It looks very reminiscent of another like system I've seen before. So we'll see. And they have a whiteboard. I love a whiteboard. Love a good whiteboard. 

Marlon Avery: Digital whiteboard. 

Sekou Doumbouya: Get that in my. If I get that, get some, some good diagrams in my doc integrated nicely. I would appreciate that. I'll play around with it. We'll see. Maybe I'll have a follow up. 

Adrian Green: Yeah, I'd be interested to see. 

Sekou Doumbouya: Yeah. 

Marlon Avery: Well guys, man, this is all this is actually. I know I don't know what you guys, but my dog was trying to say hi. I don't know about you guys, but I feel like each episode is. It's becoming more fun, authentic. It's becoming just like a flow, you know, between the big three here. 

Adrian Green: Yeah, I agree, man. Like, I was. I used to have like a. The kind of like a hum, a low hum of anxiety before these calls. Cause I'm like, oh man, like, you know, never done this before on a regular basis. Just having a conversation. I didn't know I could talk for, like to even come up with content to, you know, contribute this much. So, yeah, I'm impressed with myself as well as us with continuing to do this. 

Marlon Avery: Like, by the way, that's another shirt. The big three of AI. We gotta get that. That's it. The big. The big three events soon. 

Sekou Doumbouya: Merch. The merch store. 

Marlon Avery: Yes, the big free of AI. That's good. So guys, man, today we've discussed opening eyes. Say they are now building starting 2020, 25 up in January, they will be building computer agents. I'm sorry, computer using agents, a new software with the ability where agents be able to use and build and do things and stuff. Now without the human and stuff in the loop and stuff, you will. Nvidia, it says Nvidia now has entered into the music arena creating a new model called Fugato. And then we discuss. Uber has now elevated their gig economy to now have the ability to connect businesses with independent contractors to perform tasks for training AI models. And then over there, the new kids on the block, Blue sky has already started raising some concerns about their third party AI training. And then we, we we not Captain stuff there with Zoom says now don't look at us no longer as a video platform. We are now an AI first work platform. Addresses as such big chest out. 

Sekou Doumbouya: So. 

Marlon Avery: But now, man, it's been, it's been a great podcast over here as always. Guys, you can, you can follow us on all platforms at AI with Friends podcast on all platforms. And then you can follow me as an individual at. I am Marlon Avery on the platforms and stuff as well. Seku Adrian closest out there. 

Sekou Doumbouya: Yeah, you can. You can follow me on Tik Tok and on Twitch as the wise one. And that is number one. Yeah, that's me. 

Adrian Green: Yeah, you can follow me on Tik Tok at a green underscore Lantern with. Oh my God, is that. Search Green Lantern. You'll find it and infamous Adrian on. 

Marlon Avery: Twitter or X Adrian. We're gonna need you to pick one username too. 

Adrian Green: Oh, it's Brick. 

Marlon Avery: That is so bad. 

Adrian Green: I'm gonna, I'm gonna do that I'm gonna do that. I'm gonna pick one username. It's real scattershot, my social media, you. 

Marlon Avery: Know what I mean? You should try your best to get the official green ladder and see how much you can piss off Marvel. I don't know. I mean, see how much you put up dc. My bad. 

Adrian Green: I'm gonna try to be the UPN Jon Stewart, too, if they do that show. I'm gonna try to be the one they. They put out with upm. You know, the W. Beer man. 

Marlon Avery: Upn. Geez, I haven't heard that forever. What's that? What was. What was a Moesha. It was on the upn, man. 

Adrian Green: I'm gonna tell you. They started with Home Voice from Out of Space. I don't know if you remember that one. They started with. No, Moesha wasn't on upm, was it? 

Sekou Doumbouya: Was she? 

Marlon Avery: I feel like it was. 

Adrian Green: Girlfriends was. Oh, Moesha. 

Marlon Avery: Yeah, yeah, Moesha. Everybody hates Chris. All of us. 

Sekou Doumbouya: Yeah. 

Marlon Avery: The Parkers. 

Sekou Doumbouya: Buffy the Vampire Slayer. 

Adrian Green: And Slayer was upn. I thought it was wb. Or is that the same thing? 

Sekou Doumbouya: Yeah, that was. It was on there. 

Marlon Avery: Wow. Remember LL Cool J show in the House? 

Adrian Green: That was my favorite. I used to love that show. I watch it. 

Sekou Doumbouya: So. Enterprise. Star Trek. Enterprise, Voyager. Yes. 

Adrian Green: Yeah. 

Marlon Avery: And with top five cartoons. No, forget that. Top three cartoons of all time. There's on average as well. Okay, recess. 

Adrian Green: Don't. Yeah, you got it. You got it. 

Marlon Avery: Recess. We should reach out to them to see if we can. We can use their intro as our outro. 

Adrian Green: I'll be open to that, man. Let's see if that's, like, under Creative Commons license. Maybe it's, you know, you can just get it now. Who knows? 

Marlon Avery: Oh, wow. Smackdown was on UPN for a couple of seasons. It's always funny how we end the show. Okay, cool. All right, guys. And till next time. 

Adrian Green: See you guys. See you next time.