Space Summary
The Twitter Space PARENTING STRATEGIES TO EMPOWER YOUTHS TO RESIST SUBSTANCE ABUSE IN… hosted by ndlea_nigeria. The National Drug Law Enforcement Agency Twitter space delved into empowering parents with effective strategies to help youths resist substance abuse. Discussions emphasized the pivotal role of parental involvement, communication, and education in creating a supportive environment against drug misuse. By highlighting early intervention, parent-teen communication, recognizing signs of substance abuse, and navigating peer pressure, the space provided valuable insights and resources for families. The event underscored the importance of fostering resilience, building trust, and promoting awareness to combat substance abuse among young individuals effectively.
For more spaces, visit the Development Agency page.
Questions
Q: How does parental involvement impact youth's resistance to substance abuse?
A: Parental involvement fosters trust, communication, and a supportive environment that aids in preventing substance misuse among youths.
Q: What are the key strategies for parents to promote drug resistance in teenagers?
A: Effective communication, setting boundaries, positive reinforcement, and education on substance abuse play vital roles.
Q: Why is early intervention crucial in combating substance abuse?
A: Early intervention helps in detecting signs of substance abuse, offering timely support, and preventing escalation of drug-related issues.
Q: How can parents address peer pressure related to substance abuse?
A: Encouraging open dialogue, providing guidance, and fostering resilience in youths help them navigate peer pressure and make informed decisions.
Q: What resources are available for parents to educate themselves on substance abuse prevention?
A: Parental education programs, professional guidance, and community support services offer valuable resources for parents to empower their children against drug abuse.
Highlights
Time: 00:15:47
Parental Involvement and Substance Abuse Resistance Exploring the critical role of parents in empowering youths to resist substance abuse.
Time: 00:25:32
Effective Communication with Teens Discussing the importance of open communication and trust-building between parents and teenagers in preventing drug misuse.
Time: 00:35:19
Signs of Substance Abuse and Intervention Highlighting the significance of recognizing signs early and intervening to prevent drug-related issues.
Time: 00:45:54
Peer Pressure and Decision-Making Addressing the impact of peer influence on substance abuse and strategies for parents to support teenagers in making informed choices.
Time: 00:55:28
Parental Education Programs Showcasing available resources and programs designed to educate parents on preventing substance abuse in youths.
Key Takeaways
- Importance of parental involvement in preventing substance abuse among youths.
- Effective communication and building trust with teenagers are key in promoting resistance to drug abuse.
- Implementing structure, boundaries, and positive reinforcement at home can deter substance use.
- Educating parents on recognizing signs of substance abuse and seeking professional help when needed.
- Promoting healthy coping mechanisms and emotional support within families to prevent substance misuse.
- Addressing peer pressure and providing guidance on making informed, healthy choices.
- Encouraging open dialogues about substance abuse to create a supportive environment for youths.
- Emphasizing the role of parental influence in shaping adolescents' attitudes towards drugs.
- Recognizing the impact of early intervention and awareness in combating substance abuse among young individuals.
- Supporting parental education programs and resources for empowering families to prevent drug misuse.
Behind the Mic
Introduction and Technical Checks
Our channel it. Good morning everyone. Quick test, quick mic test. If people can hear me, just message or comment here or on our discord. If you can hear the audio. Okay, awesome. Okay, people can hear me. The ogs are replying. Music is good. Audio is good. Okay, we're just waiting for the rest of the team to join. Hey, I hear Isaac. Perfect. Is everything good in terms of my mic? Yeah, I hear you. I think. What's wrong with my Airpods? So I'm not hearing it through my earphones, but I can hear it through my phone, which is no problem. Always issues with my Airpods lately. What? Do you remember when were in Dubai? My Airpods kept acting up too. Oh yeah. Now it's my turn to. It's just not working for me anymore. My l is also here, but he just messaged me that he's feeling sick. Actually, he's been sick for a few days now, so he might just be listening and just talk when we don't know what to answer.
Community Updates and AMA Introduction
When only Mel has the answer. Yeah, but Isaac, is my audio okay since I'm talking through the phone? Okay, perfect. Yeah, everything is pretty smooth and no static or anything like that. Got it. Okay, well then I think we can start now. People are still piling in, but this spaces is going to be recorded as always, so no problem. If you're going to come late, just feel free to listen to the recording after this. So welcome everyone. Good morning, good evening, welcome to another community updates in AMA. We do this monthly and this is our August community updates in Amae. So welcome everyone. If you have any questions, please just comment it here or drop it on our telegram or on our discord. And once we get to the AMA part, we're just gonna go through it.
Monthly Recap and Updates
Okay, so the way we do this spaces is we start with a monthly recap, what's been happening since our last community updates, and then we're going to share new updates that the team has been working on and then plans for the next month. And then we're going to move on to the questions, the AMA part. So to start, let's begin with the August recap. And since our last community updates, I think the biggest update is the token burn. So we started the token burn last August 5. And as we shared before, this token burn is going to be a monthly initiative. And so far we're starting with SNS tokens that are not claimed. So on our train to earn platform, our builders and validators participate in the tasks. But after the campaigns end, they have one week to claim the rewards and after that, then they won't be able to claim rewards.
Token Burn Details
So that's the first pool that we're starting to burn. So every month at the start of the month, the first week we're going to burn whatever SNS token that haven't been claimed on the platform. Did I miss any detail in there, Isaac? No, we'll get to a little bit more details, I guess, later, maybe even during the question part, because there are not necessarily questions but just comments and sort of weird understandings or misunderstandings about token burn. So I think we should cover that. But we could go through just the updates of the month and then dig into that later. Got it. Okay. And then the second August recap is the passive rewards for Canon NFT holders. So this has been going on since the first quarter of this year.
Passive Rewards for NFT Holders
So just a quick reminder to our, all our cannon holders at the start of the month, make sure to drop by canon on exchange. And if you've staked canon nfts, make sure to claim your rewards before the month ends. Because if you don't claim your rewards before the month ends and those tokens will be moved towards the next month and basically other people can claim your reward. So don't forget to claim your passive rewards especially. Yeah, make sure to check it every month because depending on the campaigns we launch, the top words can change. So. And we post this top word usage at the first week of every month as well. Yep. Those are the two main things for August. Then we can move towards the first community update, which is new campaign capabilities.
New Campaign Capabilities
So Isaac, can you share more about what we're saying here? New types of campaigns on our train to earn platform? Yeah, for sure. So if, you know, for most of our listeners right now who have been on our platform for quite a bit, you know that the probably the most known and accessible campaigns that we've had till now are the natural language data collection. So you guys are familiar with the rephrasal of sentences. And we even did some of the word campaigns earlier on. So we had all of those campaigns running for a while. And starting this year, we've sort of ventured out. One big initiative was to onboard new clients and also support new types of data campaigns. And so we have been working on two new features on our product.
Image Tagging Feature
The first is image tagging. And it's a little, I guess image tagging is a little more simplified term. It's more robust than just image tagging. So it's like you can have an image draw bounding boxes within the image label, different parts of the image and whatnot. We have built out the infrastructure inside our product to support that. We are working with a small web three project that needs a lot of images to be tagged and labeled. We haven't started testing yet, but soon we're going to probably recruit a small amount of some of the active community members to do some testing. So that's the first. 2nd is natural language data annotations.
Natural Language Data Annotations
So when it comes to data annotations for natural language, it comes in all sorts of sizes and flavors, as say, this one that we're going to be rolling out first is in the realm of RLHF, which is reinforcement learning with human feedback. So what happens is that the client has utilized or fine-tuned an LLM to break down sentences into different parts. And we call it head relation and tail. So it's simply put, it's like you get a sentence, you identify what the main subject of the sentence is, what the action being taken on the subject or the topic is, and then what the relation is or the outcome. So this kind of campaign is almost like if you look at the validator role right now, right.
Campaign Review Process
You look at submissions by other people, and then you mark them as approve or reject. So you're sort of double checking the work. This type of campaign will work the same way where you're essentially double checking an output of an LLM. So you're going to look through these texts, you're going to see if the head relation and tail are correct. And if it's correct, you vote to approve it. If not, you reject it. Very similar concept and very similar functionalities. But this one becomes a lot more different in the fact that there's no real builders, as opposed to those sentence creation campaigns, because you're just reviewing the sentences output by the LLmdeh.
Crowdsourcing and Human Insight
This sort of RLHF data annotation campaigns are very new. And I'm not sure, maybe this is a good segue to the whole testing topic, but we've actually partnered up with Ygg to run a case study on this. The biggest case study, or the reason for case study and this experimentation is because for this whole RLHF work, what's happening right now is that AI companies are hiring people with computational linguist backgrounds. So they either, you know, graduated university with a background in linguistics or computational linguistics or any of that general field, and it does kind of get technical in the sense of language.
Challenges in Scaling AI Data Processes
So if you look at the way it's done now, it's very it's hard to scale for one company. You got to hire a lot of these very skilled workers, and then if you're going global, then you gotta do the same in all the different languages that you're trying to roll out. So what we wanted to test was, is it possible to actually crowdsource it to a, with a degree of accuracy that can be used in a commercial setting? And so we've just recently started this experiment in partnership with YGG. We are getting some of the early results in and it's looking pretty good.
Expanding Experiments within the Community
We'll have to see what the improvements are in the next iterations and so on. So that test has been going on for, I would say the last two weeks or so. And then this was just like a pure exclusive experiment with YGG. But we're going to roll out the same experiment within our own community with our current sort of high performing builders and validators. Maybe you can talk a little bit more about that. Oh yeah, definitely. You can still hear me, right? So just to jump back quickly on the big update that ISAC dropped the new campaigns capabilities.
Data Solutions for AI Companies
Now that's a huge update on our platform, because basically synthesis one, we're a data solution for AI companies. We provide them data solution. So previously what we're doing on the train turn platform was we generate raw data, which is in the form of the natural language sentences that are built discrete. But now, like Isaac said, we now have, we're building the capability to provide other types of data, like image tagging, which is huge, and text annotation. So for the image tagging, it's just, there's just so many use cases, right, Isaac, for this image tagging, like when you were talking to potential clients when were in Singapore, like some clients were building their own AI models.
Practical Applications of Image Tagging
That requires people, a lot of people, to tag a lot of images, for example, a logistics company or inventory management software. But they have like thousands of photos and they need people to tag. Oh, this part of the photo is where the food is, this part of the photo is where the products are. And you can just circle that part, basically tagging the images or highlighting what is this image. And then the other thing is text annotation, which is just providing a different kind of feedback. Instead of providing the raw data you describe what data are you seeing? And again, so many use cases, not sure what we can share.
Content Moderation Use Case
Isaac, based on the clients, potential clients that were talking to, but one was the, which stood out for me was the content moderation, right? So that's another use case of text annotation wherein we have a bunch thousands of texts and then the company needs someone to label it. Is this bullying, is this a sexual comment, is this bad behavior? Stuff like that. So basically we're building out train turn platform to be able to accommodate other data needs of companies which. Yeah, and super exciting. The big thing that you can like a telling sort of signal is that when you start to look at various use cases, there's a lot of talk, whether it's on Twitter or I just in general, but there's a lot of talk around people saying that data annotation, data tagging, labeling is going to be obsolete and that AI will do it themselves.
The Role of Humans in AI Data Processing
So to a certain degree that's true. And that there's always new tools coming out, new automation methods. AI is always improving, but at the same time, the need, the data needs get more and more complex. When you have more complex models, the data requirements to improve that model or create a next generation of that model is that much more complex. That much more requires higher quality data and whatnot. Even for image tagging, you would think, oh, like when you talk about image tagging, people are going to think about CAptCHA, like what Google or any other website does to sort of prove that you're not a bot by identifying all the images with cars or buses or whatnot.
Diverse Use Cases for Image Tagging
But then there's use cases that become a lot more specific in that one sort of person that I was talking to was at a company that does the no code web development. So I'm not sure what the best comparison is, but maybe sort of like wix or WordPress style where you don't necessarily need to have coding experience to create your website. And they were trying to build AI tool that helps with that system. But for their requirements on image tagging, they needed pretty much like web development experts, they needed people to be able to go through an image of various websites and identify all the different sort of items within the website.
Technical Expertise in AI Tagging
What are the hyperlinks, what are the hamburger menu items, what are, you know, what's the flow? And it just becomes deeper and deeper where it requires subject matter expertise. And this is a very true and appliable case when it comes to natural language text, because if you see that the current campaigns that we're experimenting with, an LLM is being used to create these data, but then we need that extra layer of human expertise to double check that all of the outcomes of the AI. So even within that whole natural language text data annotations, it becomes more technical.
The Need for Human Review
Whether it's more technical in terms of language and linguistics, or more technical in terms of domain expertise. So it's the same thing. When RPA or robotic process automation came out, you could automate away very mundane, repetitive tasks. You couldn't automate things that, like for example, if you set up an automation program and the UI for something changes, it breaks your entire automation. So then you have to reset it and reconfigure it and all of that. So we're experiencing very similar things in terms of AI data, where you can set up ways to automate, various data preprocessing.
Conclusion on Human Elements in AI Operations
But for more complex things, there's always going to be a need for that human review, the human touch and the human input, and that just creates a better model and more accuracy, and performance. Yeah, definitely 100%. And these are not just assumptions. Like this is what we're hearing I on the ground. Like when were in Singapore. So when we had the event super AI, we had this booth and our booth clearly said AI, data labeling, data annotation, data generation, and people like big companies were really going to us like, oh my God, this is exactly what we need. Because there's no, like you said, there's no off the shelf solution for technical and nuanced data labeling and annotation.
Data Marketplace in AI
One of the biggest thing I would say is a data marketplace where they say you can exchange data sets that could be used in AI training or AI model marketplaces. So these are very big in web three. But if you're actually working within software solutions with AI, you can quickly sort of realize that these aren't. It's a very backwards approach because you talk to different people in different companies, different industries, and all of their needs are so different. Of course there are underlying similarities and think commonalities that you can tackle, but it's not like you can magically find a data set that you can just use over and over in different companies within different industries and whatnot. Same thing with AI models. Like if you talk to companies, they all have different problems that they want to solve with AI. So to sort of come up with an approach where you're supplying a model first or you're supplying a dataset first. I'm sure there are cases that works, but in a lot of cases, it's just not a feasible solution.
Building Infrastructure on Demand
We're taking the exact opposite approach where we're building out the infrastructure and the tooling needed on a demand basis, on what people are coming to us with their problems for and how we can solve it, so that we have exactly the knowledge of what they need, what kind of tools would solve their issues, and build it as generally as possible in a way where we can sort of create that infrastructure and tooling to be able to apply the same tools on different types of use cases in the same sort of environment. Oh, yeah. And yeah, I think that's really our competitive advantage. We can really build those models, or we can really provide those specialized demands from potential clients.
Community Updates
So that's, so that's just the first community update, and that was a huge one. So basically we have, we're building new campaign capabilities because AI companies need them and they are willing to pay for this because there's. This is a real demand from companies. The next one, actually, this is related to what you started, Isaac, which is the testing we're doing with YGG. So we have a partnership with YGG. So YG's mission is to create opportunities for people through web three gaming. Right. And we have a long partnership history with Yggdein since. Since Quantum Oasis. So for the ogs, who still knows. What, that's something I haven't heard in a long time. When I was preparing what I was gonna say, I forgot for a bit the phrase quantum oasis that I had to search. Wait, what was the, what did they call the game before? So, yeah, YGG, long time partners of ours, and they're primarily known for gaming guilds, but they are exploring earning opportunities beyond gaming.
The Future of Work Initiative
That's why they have this initiative, future of work. So basically, they're exploring in our very tech heavy world now with web three, what does the future of work look like? And that's why we partnered with YGG, because basically our platform does provide earning opportunities and does provide basically work for our builders, validators, architects. So that's the background of this YGG partnership. And we are doing this case study with them regarding the triplet campaign, the text annotation. We're calling it triplet campaign for now. And, yeah, so far we started this, we've been talking about with them for about a month now. And we just started the tests for two weeks. And I think they're going to have an event in token 2049 and they will share, they might share the results of this case study and. Yeah, highlight, basically synthesis one in our train toward platform.
Engagement with Community and Testing
So Gabby, who's the founder of YGG, is, you know, he's a super fan of what we're doing. And one of the, as Leo mentioned, one of the biggest sort of pushes for and missions for him is to not only make YGg about gaming, but about pretty much like this. The future of work. So not only earning through playing these games and having fun, but also providing new opportunities for people around the world, for people with different backgrounds and different skill sets, to be able to tap into a bunch of earning opportunities that's not also related to gaming. And he's been partnered with us. We've had many, community members come in through YGG, whether it's YGg sea or InDGG, which I believe all of them except for YGG, have changed their branding and names. But that's another reason why we've sort of partnered up to conduct this case study, because we want to show the world that these kind of opportunities exist.
Crowdsourcing in Niche Markets
They're going to continue to come out. And for people to recognize that crowdsourcing is a viable way to scale, even in very niche or domain specific areas where you would assume or think that, oh, you need subject matter experts to tackle these, you need people, you need data scientists to tackle this. We're trying to prove to the world that's not always the case. And companies, especially small to medium sized companies, can actually take a lot of advantage and utilize the power of the crowd and really show the world the pros of doing this kind of decentralized data pre processing work. Yeah. And this triplets campaign that we're testing, it is new ones. Like, if you look at the guidelines. Yeah. It'll require you to really use your brain here. It's very different from our previous campaigns, data generation, this, you'll need to really study it.
Testing Campaign Results
But what's interesting is for the very first test with our partnership with YGGD, the very first test that we've given them. So we just gave the guidelines and then we dropped a test, 100 questions, and so far we're seeing upwards of 70% of accuracy from crowdwork. Yeah, it's super significant because, you know, earlier I mentioned that the requirement for someone to get hired into a role that's literally doing this job is to have, you know, whether it's a college degree or some kind of specialization in computational linguistics. And for someone without, like, for like a random sort of a person without that background, let's say they just loved to read and to sort of just have a good understanding of language to perform even. I mean, 70% to some people might seem like it's not good, but you just have to think about all of the requirements.
Significance of Background Knowledge
Right. Like, it's not an easy task. It's. It requires all this background for a reason. And for someone without that background to do that on their first try is very impressive. Yeah. There wasn't no specialized training. We literally gave them a document of guidelines and read it. So imagine more if there's more feedback or more communication, more training, I think that accuracy rate will just increase. So, so far promising. And then, yeah, I think this is something that we can write an article about as well on our side and share with the community after the testing is done. Yeah, for sure. Which. Where the target now is by mid September, I think we should have the results. And yeah, we can definitely share it with the rest of the community.
Upcoming Community Testing
So for this campaign testing, so we talked about YGG is doing, but our own community will also be doing it really soon. So we've made announcements on our Discord top builders. Basically our level four builders were invited to try out a new type of campaign. And so if you're listening to this spaces and you're a level four builder, or you know that you're a really good builder and you've been building for a time now with us, feel free to open a ticket on discord, drop your wallet address and tag me Leo. Leo 6670. That's my, that's my discord id. So make sure to create a ticket, drop your wallet, tag me on discord, and we're gonna give you the pilot tester row.
Community Engagement and Participation
So those that have already done that, those that have reached out to me, they. We already gave you the guidelines and a sample test. And right after this AMA or tomorrow, I'm gonna drop, we're gonna give you the first set of tests to try this new triplets campaign. So yeah, if you're an OG, a top builder, a good performer, just open a ticket and we'll give you the role. Yeah. Exciting, exciting times ahead. Yeah. And it's. I mean, if you're into like.
The Potential of AI Campaigns
Aih. Hey, everyone. Just checking. Can you still hear me or you can hear Isaac, but not me? I think we lost Isaac. Oh, there we go. I hear you. Got it. Okay, you're back. Well, I think I was right. I was saying something as I got disconnected, but I forgot because I got this. We were talking about, were talking about, oh, how our community will be able to join the beta or are joining the beta now. So I called for stuff builders, ogs. Oh, you're saying something like, if you're an AI. So if you're, if you're at all interested in the whole AI space, I think these new types of campaigns will be super interesting because it's really at the very bleeding edge of AI.
Developments in AI
meaning that, you saw the sort of, how LLMs grew and got their hype and got developed in a super. Well, it got developed over a long period of time. But like, you see the recent progression, how like, good it's gotten from chat GPT to, you know, GPT four. Oh, and all these other open source models coming out as well. But that being said, LLM still suffer from hallucinations. They still are glorified, next word, predictors. So they don't necessarily understand actual what you're talking about. And so that's why when it comes to actual enterprise deployments, you don't see LLMs being used in customer facing settings or very mission critical type settings.
Improvements in AI Applications
And one improvement that's been done in many places is what's called rag, which is retrieval augmented generation. So it's sort of narrowing down, I mean, simply put, it's sort of narrowing down the knowledge base that generates the queries and answers. And even within rag, there's different types. And one field is, or one option is knowledge graph based rag and stats. It improves the current LLMs by a certain degree, but at the same time, it still doesn't encode logic. And to be able to perform logic is to think more like a human being. And so what these triplet campaigns, these new data annotation campaigns are doing is we're helping pretty much encode logic into natural language text.
Advancements in Wisdom Graphs
So I'm not sure if anyone is familiar with wisdom graphs. So wisdom graphs would be magnitudes sort of higher than a knowledge graph. That's sort of the frontier that we're working in. And by contributing these to, by contributing to the campaigns, like, you're literally working on things that are being applied at the bleeding edge of AI. Yeah, definitely. These are things that, yeah, people don't talk about because we're still building it, but this will be the future. When you were saying it, I kind of remember the comment we're in, that it's like LLM on steroids or basically just what will replace LLMs in the future.
AGI and Future of AI
Yeah, it's definitely like people throw around AGI or artificial general intelligence or ASI, artificial superintelligence. But you know, I'm on the fence of, I'm on the side of, you know, we're never going to get there until with just LLMs because you know, it's just a prediction algorithm. Yeah. Awesome, awesome updates. So quick recap. We're building new campaign capabilities. We're into a partnership with YGG. Our community will is testing the new type of campaign as well. And yeah, this type of campaigns, they're not your usual, these are not your typical, I'd say, boring campaigns.
Community Activities and Events
These are on the leading edge of AI development. Super exciting stuff. The next community update is supposed to be we're going to talk about a new Zealie and Galaxy quests. These are for community activities and we're going to have rewards for our current active members. But Mael isn't here to discuss this with us so we're just going to drop it as announcement on our discord server and telegram over the weekend. So watch out for announcement this weekend. Basically then the next community updates is our upcoming events. So September will be very busy for everyone.
September Events in Focus
For September we're just going to continue marketing, attending events, building partnerships, just spreading the word about Synesis one. You want to start this Isaac, or you want to list on the event? I could briefly talk about it. So September is pretty big for not just us but the whole crypto space as a whole. Arguably the largest crypto event is in September, which is token 2049. So starting early September I'll be flying out to Korea for Korea Blockchain Week. I am sort of in the process of finalizing and securing a, a speaker spot at a AI related event at KBW. And then right after Korea Blockchain week I'll be flying over to Singapore for Token 2049 and Solana Breakpoint.
Token 2049 and Beyond
there we'll have not just myself but our CTO Leo and our co founder as well. So the majority of the team will be in Singapore for Token 2049. There during token and breakpoint week I will be speaking at a side event. I think it's called multi Chain Day. It'll be announced soon as well, but it's the same side event that I was at for consensus in Austin, Texas earlier this year. So it should be a really great event with pretty good turnout, but this time we're going multi chain, I guess, not just Solana focused. So, so far.
Collaboration with YGG
So there's two of those panel speaking slots. And then also were talking about running the case study with YGG. They just recently notified us that they'll have a booth at Token 2049, and then they want to showcase sort of the what's been happening on that feature of work front as well. So we'll be working with them to show some of the results, talk about some of the things that we're working on together and whatnot. So we should have some good exposure at both Korea blockchain week as well as token 2049 and then breakpoint.
Team Structure and Upcoming Events
Did I miss anything? Leo? No, that's it. After breakpoint, the team will be coming together as well for our teamwork. Yeah. So for people who aren't familiar with it, we are a fully decentralized team. So we have people everywhere. Like, I'm in California, Leo's in Manila. We have developers in Turkey, Korea, and then, you know, everyone's everywhere. So we are a fully remote team, but we try to get together in person maybe two times a year. And during this time, everyone sort of spends a good period of time together working in the same place physically. This helps with a lot of the ideation as well as alignment in terms of the next half year to long term vision. So we've done this earlier this year in Dubai, and that's been pretty fruitful. And the next one will be right after breakpoint. So right after breakpoint will be, will continue to be in Asia and then just working away together and setting the short term and the midterm goals for post breakpoint.
Event Participation and Communication
Yep. Yeah. So that's it for the upcoming events. Korea blockchain Week and then token 2049 and then breakpoint. And we'll be, we're trying to secure speaker spots and we'll be at AI panels and some of them. So exciting. September. We'll keep everyone posted. We'll be sharing what's happening on our Twitter discord and telegram as well. I think that's it for the project updates. Isaac, did I miss anything before we jump into Q and A? No, I think. I think that's it. Yeah. Huge, huge updates. And we're gonna have all this written down because by the end of September, that's time for our quarterly newsletter as well. So all these updates, we're gonna definitely put it into an article so that everyone can read about it as well. Okay, let's jump into the fun part. Let's jump into Q and A for everyone listening. Feel free to comment, to reply your question on the street or drop it on our discord.
Addressing User Questions
Oh, we have a bunch of questions lined up already. Let me set it up. Oh, I lost the channel. Questions, feedback? Here you go. Oh, first, 1st question. Isaac. This is a new one. What's the plan on getting listed on centralized exchanges? What? Oh, the plan for when listing. Listing. So centralized exchanges is like very finicky. Like it's, unless you work in the crypto space and actually deal with decentralized exchanges, it's hard to understand. A lot of people think it's, you know, you just say, okay, I'm going to list on this exchange, and then list. But it's a little more than that. One thing is market conditions. The overall attitude and negotiations change based on market conditions. When it's a bear market, everyone is very slow. Centralized exchanges don't want to list tokens at the moment. And then as you start to get into a bull cycle, they want to list like literally everything, and charge tons and tons of money to list these tokens.
Strategizing Exchange Listings
And you know, that being said, like, we are always open to centralized exchange listings, but once you list, on a centralized exchange, it costs a lot of money. Like the listing fee itself costs a lot of money. And then the market making to maintain token activity within the centralized exchange costs a lot of money on a recurring basis. So you have to really weigh the options. Like is it more fruitful to pursue a listing in, let's say, a tier two or tier three exchange and pay all this money, or is it better to spend that money elsewhere? And whether it's on marketing or product development. So you have to sort of weigh these things through. And that's why you haven't seen us listing on centralized exchanges as much. We did talk to, you know, all of the ones that always get brought up in our chat, whether it's Mexc, gate, Kucoin, whatnot.
Market Strategy and Client Engagement
Like negotiations always change. We always have that open door to these guys. But if we're not listing yet, that means it doesn't make sense, whether it's a financial or timing, but we are being strategic about it, and it's like a double, both pro and the con, right? To a certain degree, you do get more liquidity and more exposure from decentralized exchange listings, but it's not the same as before. It's not that effect isn't as strong in today's. Day and age back, maybe even back a year or so, arguably, to a certain degree. Yeah. So you really got to think about it. And we're always open to, if it's like a tier one exchange, then definitely we'll pursue that. But when it comes to tier two and three, it's. We really have to think carefully.
Plans for Marketing and Client Acquisition
Yeah, definitely. Next question. Is there a plan for marketing? I don't know if we have a marketing budget. Yeah. So marketing. So we've tried many different things. Right. And then, so the big focus for us this year has been growing out the client base, and that's how we got to a stage where we're rolling out two new features to handle two new, different types of these data campaigns. And with that goal of getting new clients comes a change in terms of how we do our marketing. And one of the ways that we've been spending our marketing budget is to participate in these different industry events. One big one was super AI. We were a sponsor there, and it was a great AI conference to be a part of.
Industry Engagement and Company Growth
And I think we made a lot of progress there in terms of getting a lot of new hot leads and just sort of being able to gauge where the industry is, what people are looking for, and what's happening. So we're going to continue on that route. We're going to be sort of allocating a lot of budget towards attending events, having our boots on the ground, getting more media exposure, and getting more, rubbing shoulders with a lot of the other leaders within the AI space, not just AI crypto space, but then the AI space as a whole. And that's sort of our current strategy, and that's where we're spending our marketing money right now. Yep. The next question is, where are we at with attracting new customers?
Client Development and Project Opportunities
So this question I think Isaac just answered through the marketing question, where we're attending events, we're getting hot leads. And this new train to earn campaign feature, which is image tagging, was actually, sparked by, potential client, but it is also going to be usable for other clients, other use cases that will require image tagging. So, yeah, we are working on getting new clients. Yeah. And it's much easier for us to work with smaller projects and companies because when you deal with larger companies, like, they need a solution immediately. And for us right now, we support a few. We have the data collection, text data collection campaign, and then we're rolling out the text, data annotation and then the image labeling.
Expanding Solutions for Client Needs
So until we get to a comfortable suite of solutions that we can apply to different domains. It'll be hard to service bigger clients. So we are working with a few smaller perspectives and that way we are being able to get work very closely with those teams to understand their requirements and then build out the right solutions. And that's how you're getting these new types of campaigns. So it's all underway. Some of the bigger ones that we're talking about take a long time. I mean, even medium sized companies, it takes a long time. These deals don't happen in the matter of weeks. It's more on the matter of months. Most of the time before they engage in a deal, they want to see a POC first.
Addressing Market Conditions and Investor Confidence
And then you got to run through all these different hurdles per se until they're willing to actually pay for it. It's just typical software sales. But the good news is that we're getting a good stream of these new projects that are looking for these solutions. Next one is the team taking price slide into consideration? What are the plans to address this as we're moving to lower despite general market sentiment picking up? Yeah, I think, I mean, obviously, like our, the team's incentives are aligned with any token holders. Like, it's not like we're adversely incentivized. It makes no sense. Like, obviously, the more the token appreciates, the more, the team earns as well. So that is definitely, a focus.
Long-Term Vision and Business Development
But, it's the same, it's the same sort of question and argument and, you know, discussions that happen over and over. Every time there's a dip or a pump, it's. So when there's a down cycle, it doesn't make sense for whether it's a project or anyone to really fight against the market. One company cannot really fight the wave of an overall market condition. In the same way, if you're just spending money to pump your own tokens that, like, that doesn't really, like, it's not a real project. You're not like, we choose to spend our money on developing products, unlike a lot of different other companies that prioritize on, let's say, marketing over having nothing.
Focusing on Product Development
So it's. I think it's. I'm not necessarily saying that's bad, but if that's the kind of project that, you know, you want to invest in, then there's plenty of those. Like, what we're doing is we're trying to build a real business with real products and real clients. And that's, you know, if that's something that you like, then I, obviously, it takes time, and that's what our priority is. That's where we're spending our money to build a real business. And we see this as a long term thing. So I know a lot of people investing in crypto are very short term investors.
Team Commitment and Market Challenges
So I think it, the best thing we can say is, yes, we are aligned in the incentive of token price, and we see this as a long term thing. It's not like we've completely abandoned something. It's still our livelihoods. Definitely. On that note, I also want to address, there's always people saying, oh, we're dipping because the team is selling. It's like, no, it's like, why would we sell at such a low price? We, we've, like, if anything, we've put in so much work on our end that we would be selling ourselves short by selling, because we know that the value of all the effort, the blood, sweat and tears that we put into it is, you know, much greater than what's reflected right now.
Addressing Market Perception
So, like, I don't know where, like, they just randomly say these things. I'm like, okay. Like, it makes no sense. So it's. Yeah, it happens every cycle. Oh, yeah. When it's get, when things get really tough. Yeah, this kind of things just pop up. Another question from crownie. Are AI companies paying for, are AI companies paying us for the data work in SNS? And if so, is it helping us to shore up on the token price? So there mainly not paying in SNS, they're paying in fiat or stables.
Revenue and Strategic Decisions
And so that goes towards buying the tokens needed to put the budget in for the campaigns. I think this is one of the big strategic decisions that we made because we don't want to sort of turn off the web two companies by saying, hey, we're only taking project tokens. Well, if they already have sns, they can just set up the campaigns on their own. But when we work closely with them, they'll pay us in whatever payment method is best, and then we'll use that to buy SNS for the campaigns. So as we grow, that will create bigger and bigger impact and buying pressure on the token. But right now it is a small company, and although we are making revenue, it's not like we're making tons and tons of revenue.
Future Growth and Expansion Plans
So it's more of a growth thing where you'll start to see the impact as we grow bigger and bigger. And is the business expanding at all to more AI companies in need of data? Yes. And that's exactly what you're seeing with these new campaigns that we're rolling out. And, yeah, it's a process. Yep. Next question is about liquidity.
Increasing SNS Liquidity
I was told that there were plans to increase SNS liquidity on chain. Where are we with this right now? Yeah, so starting from the late part of last year, when there was a big boom in AI crypto side, there was a lot of concerns about the liquidity on the Dexs. And from that point until starting then, we started adding liquidity into dexs. And went from, I think it was probably maybe less than one hundred k of TVL for liquidity to over a million. And so, of course, there's outside people putting in liquidity into the LP's, but majority of that initial increase was from us. So that it's been there, we've done that. Obviously, we'll continue to do more as time comes and when the token price drops and when the market has a downturn, all of those metrics go down with it, how it works. So, you know, it's. It's something that we've already been doing and we will continue to do.
Future Campaigns and Roadmap
Next one is with the introduction of new campaigns on train to earn. It seems we've addressed the 2024 roadmap. What can we look forward to next? I think it's continuing to grow in terms of clientele because the more features we build out within the product, that means we can start to serve more companies that have similar needs. So by doing this, we're setting ourselves up to scale better. And that's going to be a continuous, big focus, because we are again building something that turns into a very good business. Yeah. Next question from arcade. For current campaigns, we have builders and validators to check the data, which makes perfect sense. But how do the triplets get verified once they have been worked on? So this is very interesting, arcade man, because the current campaigns, it's all human generated, right? The builders generate the data, and then validators will check the data that our human builders created. But for triplets, it's quite interesting because we're gonna start with model generated data.
Verifying Triplets Data Generation
So our client has created this model that generates this type of triplets data. And it's our job, the human and the loop, our workers, to check if the data generated by the machine is correct or not. So this is what Isaac was saying about RLHF, which is the reinforced learning with human feedback. I suddenly felt sleepy mid sentence. 01:00 a.m. i just stopped talking. But yeah, this is very interesting because this is where human in the loop gets valued in because we start with machine generated data and then humans need to verify, oh, was the data generated correct or wrong? And the way we've set it up so far, it's going to be consensus of voting. So for example, ten people will vote on this sentence in this triplet that the AI model generated, and then we're going to see, well, what's the voting? Did the majority say that's correct?
Majority Voting and Feedback
If the majority voted correct, then we will accept that the model was correct. But if the majority of the users voted reject it, that means then we're going to provide that feedback to the client saying, hey, people, don’t think that the triplet generated by your models was wrong? Then they’ll review what went wrong and they're going to fine-tune it. So basically, what we're doing is we're helping fine-tune this model created by the client. Did I miss anything there, Isaac? No, I think arcade man brings up a great point and I think this kind of thinking is very helpful because it's something we have to figure out as we see the results. So as Leo mentioned in currently, we're trying to see that if. We're trying to see if it's possible to do on a very accurate basis just using consensus of voting, but it might not. It might work. So we'll have to see and adjust from there.
Burning vs Buying Back SNS Tokens
Yeah, yeah. That's why we're doing the testing with YgG as well as our top builders so that we can improve this type of campaign and see what's possible. Okay, I think that's it for the questions on discord. But I also see on discord, like also why not? Why not buy back sns from the market instead of burning it? So I think this is. Yeah, and let's talk about burning. A lot of misunderstandings about burning. So to address, why not buy back sns from the market instead of burning it? We're burning the SNS that's gone unclaimed from the rewards. So the campaign budgets are set, people do the work and then they have to claim the rewards. So if you don't claim it, then it goes to a burn wallet that we then burn.
Understanding Token Burn Impact
So this is completely different. We're not just doing buy and burn, which I think could be a big possibility later on. But our first implementation of token burn is to burn unclaimed rewards. With that, there's just a weird outlandish statements, like someone’s like saying that token burn is leading to a price decrease and whatnot. So token burn in itself has no effect on the price by itself, right. All we're doing is we're taking a certain portion of tokens that are in a wallet and then burning that, meaning it's taken out of total supply and circulation. That in itself will have no impact. We're not going to an exchange and selling or buying tokens. So that in itself has no effect. The way it has effect is if the company valuation is let's say $100 for 100 tokens, then you burn 50% of that token.
Company Valuation and Token Prices
Then technically if the company value itself stays the same, then the token should be worth twice as much because the supply is reduced by half. Nothing has changed within the company. It's still valued at dollar 100. But the number of total supply of tokens change. So then the price has to now reflect that. So then that's where the whole price effect of burn happens. And so burning itself doesn't mean that we're buying or selling, it's just taking supply out. And I just want, I don't know, like I don't know where they're getting these random information from. Yeah, hopefully it clears it up. Hopefully. And then we also have questions on our Twitter, what's the official amount of tokens in circulation? So for this one they can go to coin market cap, right.
Token Circulation and Information Sources
That one is verified circulation supply. Yeah, you could go to like Birdeye, so as well. Or Coingecko. It should have, it should have the, well I guess bird eye doesn't have circulating, but total supply is fairly accurate I think. Let me think. Yeah, Coinmarketcap, coingecko. They both should be pretty reliable. I know one of them doesn't update as quickly as the other. I can't recall off the top of my head, but we've given them APIs that should automatically update. We've done our job because they asked for it and they should be updating it however frequently that they should. But one of them is slower than the other for some reason.
Token Tracking and Updates
I just checked and based on the APIs that we gave, it's coingecko. That's the same with what we give. Okay, so Coinmarketcap is the one lagging in their updates. So yeah, Coingecko, check coingecko for circulating supply. Next one is oh these are the same questions about getting new clients. Oh here, there's a scam out there airdropping tons of asns using our logo. Any words or comment on this? Yeah, so I also get those weird fake tokens. Please don't click on anything any links or anything like that, those are fake scam tokens.
Caution Against Scams
And unfortunately this is the web three space, this is blockchain space. There's nothing we can do about it in that there's no central authority where we go and be like, hey, can you take them down? Because anyone can create any new token on chain and then just name it whatever they want, send it to whoever they want. So this is the negative side of being in this sort of space where everything is permissionless and there's no governing authority. So that's why like people just need to be very careful. Yes, please be careful with your funds, your wallets, where you connect your wallets. Always verify who you're talking with and the links just don't just blindly open links that you got.
Passive Income Generation from NFTs
DM’s. Yep. Be careful out there. Are the NFTs going to benefit in terms of passive income generation from the new type of campaigns? Interesting, because. Yeah, go ahead. I mean, short answers. Yeah. The long answer is once they become campaigns that are running on our platform, then it becomes, you know, it just programmed in. So. Yeah, because remember, our passive income rewards are funded by a portion of campaign fees.
Funding Passive Income through Campaigns
So every time we create a campaign on the platform, there's a platform fee and a portion of that goes into funding the passive income. So any type of campaign, basically that we upload on our train toward platform that will benefit the passive income pool. Are you last, the last question I see here is are you thinking of any different ways to add more utility for SNS besides taking. Oh yeah, that's something that I forgot to cover. But besides everything that we talked about, there are a few other pretty big things that we're working on behind the scenes.
Upcoming Partnerships and Utility Enhancements
It's too early to reveal at the moment, but I will say that there's a partnership involved with hardware, so I'll leave it at that. But it's with these new sort of partnerships and collaborations, we are always thinking about ways to bring more demand and more utility to the token. And I think this one will be pretty good. So I mean, by the time we're done with token 2049, it should be revealed. Oh, another surprise. Maybe by the time of our next community updates. Yeah, so it's pretty big. It's pretty big.
Community Updates and Ticker Changes
Yep. So our next phases might because we usually do this community updates and Ama mid month, but given all the events in September, we might push it to third or fourth week of September, just heads up to our community and then last one, could we change or ticker. Too many. Too many sns out there. What should we change it to? What's your suggestion next year? Yeah, maybe. I don’t know. Maybe that's our next campaign. Suggest new tickers.
Challenges in Changing Tickers
Yeah. I think any ticker we change, there's going to be, I don't know, all the tickers. It's like a never ending loop. Right. Also, because then anyone else could just be like, okay, I'm going to create another one that's the same ticker. So then it's like it just playing, you know, rat and what? Mouse and cat.
Conclusion of the Discussion
Cat and mouse. Cat and mouse. Where you change it, someone more people create it, you change it again. Yeah, it’s just a rat and mouse game. Okay, I think that's it for the questions. I see. Are you seeing anything else? Let's see. I don't think so. Yeah, I think that's it for the questions. If we miss anything, or if you suddenly you remember any other question, just feel free to drop it on our telegram and discord. And, yeah, we won't wait for the next AMA. We'll just answer it there.
Closing Remarks
So I think that's it for our spaces, Isaac. Awesome. Thanks, everyone. And, yeah, as always, it's recorded, so you can always catch up. And, yeah, just be on the lookout for some of the more exciting stuff coming out next month. And if you'll be at Korea blockchain week, token or breakpoint, let us know. Happy duties with you guys.
