Decentrally Validated Simulation of Automated Trading in Artificially Intelligent Markets: Risk-Averse Agent Optimization
You can soon partake in a powerful decentralized simulation by developing trading algorithms in any language and enter them into the taos.im token-intentivized races (refer to README). In doing so, you can take advantage of deep limit-order-book (dLOB) data, and with limitations to be relaxed later on, develop deep-intelligent high-frequency trading (dHFT) type of faster strategies.
On TAOS, the specific task is to create reward-risk maximizing algorithms that consistently outperform the other agents. These token-incentivized races operate 24/7 like real crypto or FX markets, with many race-simulations run in parallel to maximize the statistical significance of the average. The best race-performers are likely to be attractive candidates at real-world exchanges. The decentralized TAOS simulation applies open-source stylized agents combined with complex and intelligent AI agents, producing increasingly realistic dLOB data. This, in turn, allows traders (later even dHFTrs), exchanges, and market regulators to run more refined experiments on their own or upon request in a economically highly efficient manner, leading to better market quality.
The team has years of experience in high-frequency data recording, AI, and its implementation in market-making simulations. Now, we are taking certain feats to the forefront of decentralized AI:
[C++ CODE] The developed subnet validator logic in C++ is special in the Bittensor ecosystem; C++ code is applied to achieve the maximum level of efficiency and number of simultaneous runs.
[RESEARCH] The team includes a high-frequency data expert with 20+ years of experience in agent-based modeling, trading, and market microstructure, and others with a doctorate degree.
[HF-DATA] The simulations allow for L3 market-by-order data to be used by the AI-Agents, motivated by the fact that deep-level HF-data may be useful in AI prediction (refer to EVIDENCE).
For more info and requests on AI trading and/or the simulation opportunities, email us at to@taos.im. You can also partake in running the framework by staking and earning dividends; for info on investments, email us at stake@taos.im, read Bittensor docs, or use "CQ" to do it yourself.
