Brains of Crypto: Discussing AI and NFTs with Seagen's Team
Navigate the NFT market with Seagen's Seafarer app. Discover how AI powers price predictions and meet the crew steering the future of crypto trading.
In the vast digital ocean of OpenSea, where virtual treasures await the daring, sails a valiant navigator, Seagen, guiding the Seafarer app through the waves of blockchain. With a daring spirit and a vessel powered by AI's keen insight, they embark on a quest to unlock the hidden realms of NFT riches.
In this special edition of Brains of Crypto, we delve into a dialogue with Seagen, exploring the innovative currents that drive Seafarer amidst the broader crypto ecosystem. We aim to shed light on the Seafarer app's approach to navigating the OpenSea, and the spirit that propels this digital voyage forward.
Let’s set the sail and dive into the story of Seagen!
Please introduce us who is Captain Cuck and the Seagen crew?
Captain cuck is the meme-coin branding we originally went with to put a friendly face behind our AI utility. We have since begun to migrate away from that branding and are focusing on “Seagen”. Which is a play on the word “Degen”, the pun is thematic to our SeaFarer AI for navigating NFTs.
The Seagen crew are a group of passionate developers and web3 builders all based in Ireland. The team consists of:
- Ted Stumbrs
- David Hynes
And some other developers who we have worked with in the past, all based in Ireland.
Ted - Co-Founder. Leads the project as the main developer on all fronts (AI, front-end, solidity, back-end), and most recently, has become the face of the project. The idea for Seagen has been a culmination of his vision for future web3 trading incorporating AI technologies. All developments thus far, and in the future are oriented around market demands.
Dave - Co-Founder. Lead administrator and graphic designer. Dave is responsible for creating strategic web3 partnerships with other dApps that have a mutual benefit. Dave is a talented graphic designer and is responsible for the majority of graphics we use within our project. Ted and Dave have been working together and have known each other outside of web3 for over a decade, their most recent project: https://novablox.ai, was a successful venture that got the support of the Fantom Foundation, ultimately.
Crisps - Co-Founder. Lead marketing strategist and partnerships coordinator. Crisps has an outstanding skill for networking and is highly efficient in his approach to creating partnerships and organizing AMAs and twitter spaces. Crisps was a part of the Novablox team towards the last few months of their activity. Crisps has vast experience leading projects in web3, even with his own highly successful NFT collections on Fantom, such as; Super Happy Fun Bears (SHFB), which saw great results in both the original and the sequel to it.
The narrative sets a whimsical and adventurous tone. How does this tone translate into the user experience of Seafarer?
We went with a more adventurous branding as it does two things: It instills a sense of excitement into the project and the unique, whimsical tone we’ve set also attracts people looking to find new utilities without being bombarded by technical jargon. Our website has a theme which resembles the wording we use to brand the project, wherein there is an emphasis on oceanic, exploratory and water-based tones.
How does Seafarer app utilize AI?
The genesis product that we have released on the 28th of September 2023, incorporates a series of machine-learning techniques to make predictions on the price movement of an NFT collection.
The Technical: Currently we have down-graded to a Random-Forest Classifier that runs at an accuracy of 82% but will soon be upgrading that to an LSTM (Neural network) once we have the resources to provide a competent infrastructure for it. Long Short Term Memory Neural Networks are highly efficient at predicting time-series data. However, they require a static input size on the amount of transactions that you pass to them. Given that we are working with transactions over a 24 hour period, we cannot say for certain how many there will be and would prefer to avoid padding or reducing data. We are using a number of open-source API tools to garner data on NFTs, including all metrics acquired from an undefined size batch of transactions. This data is actually available to the end-user specifically for making a more informed decision on a trade.
The Concise: We get all NFT data per each collection in the top 50 NFTs by volume on request, then a user can make predictions via our AI model to figure out whether or not a NFT will go up or down in floor price. We also get a bunch of data from each NFT collection to display to the end-user, so they can make a more informed decision on their next trade. Note, the AI model prediction should never be taken as paramount. Always do your own research before making a trade.
How does the AI technology in Seafarer app ensure the accuracy and relevancy of the NFTs it identifies for users?
For starters, the AI only works on the top 50 NFTs by volume over time on OpenSea for the time being, this is to make sure we are not displaying any scam NFTs to the community and because these particular NFTs have vast differences in floor price + enough data for them to be used on our AI model. Essentially, most of these NFTs have been around the block for a long time, so our model can make more accurate estimations on how they will proceed next. We can never ever guarantee 100% accuracy for any price prediction but we can always give people insights with our AI.
The AI doesn’t monitor relevancy, however it does display total transactions for any given period. This, in some sense, tells a user how popular an NFT is for that time period.
What were the main challenges encountered while developing the AI technology for Seafarer?
The number one, and biggest challenge thus far was identifying what metrics to use for data collection. Of course if we simply focused on one data point: for example, Floor price over time, we would surely be able to create an accurate model with binary labeling, but given the nuanced nature of Market activity, there are always more factors at play. So we decided to utilize other metrics for NFT prediction, such as, Floor price, Floor price over time (as a coefficient), Volume and Volume over time (as a coefficient).
The second biggest challenge has been handling the infrastructure, we had to ensure that our model would not be compromised and that it can efficiently post data to the front-end-user at an optimal pace and with minimal errors stemming from edge-cases. We actually have built an API to facilitate this massive computational burden. Which we will eventually release as a product of its own, so that developers in the space can make calls to our AI seamlessly within their own environments.
How does your governance model empower the Seagen community?
We don’t currently have a token governance model in place, however most of our project decisions are run by the community so that they are made aware of any critical changes and can offer insight into how to proceed. We always take on board what the community has to say and apply changes based on their needs and demands.
What are your plans in terms of growing the community?
We have a growing network of people on Twitter and Telegram who have chosen to support us by hosting our founders in their lounges and daily twitter spaces, for starters.
We have dedicated a significant portion of our funding to financing community growth incentives.
For which, we have already done: buy-competitions, giveaways, invite contests, many AMAs on our own platforms and as guests on other people’s platforms.
We hope to have released all of our future developments by the end of October. This includes an ERC20 version of the SeaFarer AI. Once these products are out there, we believe that they will speak for themselves, and that seagen.app will be a necessary stop for any trader looking to learn about their ERC20 tokens or NFT projects.
What are the future plans for Seafarer app in terms of features, partnerships or expansions?
As we have mentioned above, we plan to release an ERC20 version of our price prediction AI before the end of October. This will come as a token suite that provides insights into the safety of tokens (whether they are honeypots or not), will give some transaction data to the end-user for each token trade on the open market (mempool data). It will also serve as an advertising point for other ERC20 projects.
From these transactions a user will be able to gain insight into the price movement of a project (graphs) and based on the trends that our AI picks up, they will be able to get a prediction on the price of that token in the nearest future.
For our NFT AI, SeaFarer, we plan to release a browser extension version of it, that will operate on the OpenSea marketplace seamlessly and provide insights into tokens while a user is scrolling through the pages of collections.
What are your thoughts on the current state of the NFT market and where do you see it heading in the near future?
We believe that the NFT market directly follows the overall crypto market. The current downturn of volume and activity is based entirely on the fact that we are going through a bear market. However, we have seen many projects sell out even through the bear because of their strong community support. This is indicative of the fact that there are still militant NFT-heads out there looking to grow their collections or flip a quick profit. That being said, once BTC and the overall market returns to its former glory, we will most certainly see an up-take in NFT activity, which would be great news for our dApp!
What are the most significant challenges facing the NFT market today and how can they be addressed?
The biggest challenge to getting into NFTs is the barrier to entry.
We believe that getting into NFTs isn’t as quick and clear a process as one would like.
One must learn a lot of new terminology and technology before getting into any NFT collecting. Tools like SeaFarer make that experience significantly less daunting and more beginner-friendly.
However, there must be a focus by other dApps in the space to create a seamless, beginner experience for NFT trading, accessible to everybody!
Once that happens, we believe that NFTs can be adopted en masse, even outside of web3 technologies.