🤖The Factory

The Factory is the research and development hub of Block by Block, where three artificial intelligence (AI) models are actively gathering data and developing their capabilities. These models are designed to work in concert, each contributing its unique skill set to provide an exhaustive analysis of the memecoin market.

NYOTA (Name Yield Outcome Token Analyzer):

NYOTA is inspired by Nyota Uhura, the renowned communications officer and linguist expert. Its role involves translating token names and keywords into embeddings using BERT modelling. This capability allows NYOTA to identify current trends, narratives, or 'metas' and potentially spot emerging themes. Notably, NYOTA can cluster similar token names together, and will soon be able to provide users with a quick overview of average returns for any given trend. NYOTA's analysis will also be incorporated into Alerts where, in addition to setting filters for individual keywords, you will be able to add search terms for trends e.g search for "Chinese Meta" to be alerted when any token matching that filter is launched.

More importantly, NYOTA is also designed to feed Professor X with these rich, nuanced embeddings, enhancing the mastermind's analytical power.

Each dot represents a different token name, NYOTA is able to characterise and cluster these token names to identify past and present trends and potentially spot emerging narratives early

SHERLOCK (Scam Highlighter and Early Recognition of Legitimate Operations through Contract Knowledge):

SHERLOCK's mission is to identify and catalog scams. With approximately 600 new tokens launched daily, and well over 50% of those being scams, SHERLOCK is designed to dissect these scams and reveal the tricks scammers use to dupe average investors. SHERLOCK specializes in uncovering the less obvious scams that use renounced contracts and other similar tactics. Users can contribute to this database by using the command "/scam_contract" followed by the contract address. Soon, SHERLOCK will provide a scam probability score for any entered contract address and draw connections to similar contracts potentially launched by the same developers. Importantly, SHERLOCK's insights significantly enhance Professor X's analysis.

Professor X:

Professor X is the orchestrating AI. Harnessing the data from NYOTA, SHERLOCK, and The Hive, it meticulously examines each new ERC-20 token, capturing multiple snapshots and collecting around 50 input features per snapshot. These features range from token-specific details like market cap, taxes, liquidity status, deployer balance, to environmental factors such as gas fees, day and time, and the latest values from Block by Block's Fear and Greed Index. Professor X even scrutinizes holder information, the token's social media presence, website, whitepaper, and any linked influencers' performance data. With scam probability data from SHERLOCK and a thorough review of the token name by NYOTA, Professor X can build an incredibly detailed picture of each token.

Currently, NYOTA and Professor X each have around 3,200 records (representing around 150,000 data points), and data collection for SHERLOCK has just begun. Collection is temporarily paused as new features such as Twitter functionality are being integrated. After this integration, the raw data will be made available on the dApp. Once NYOTA and Professor X reach 20,000 records and SHERLOCK hits 5,000, the data will be preprocessed, and deep learning/machine learning models will commence training.

Eventually, when Professor X can reliably predict a token's launch shortly after it happens, its signals will guide The Treasury's investments. Once consistently profitable, these investment results will be made available to users, and Professor X signals will be incorporated into Alerts. Users will then have the option to apply custom filters such as risk profiles to further refine these signals. And Professor X will be renamed Professor X’s 😎