Artificial Intelligence for Trading: Insights from My Experience and Results

Discussion in 'Trading' started by joaco7172, May 11, 2024.

  1. AI is massively overrated. The only real tasks AI can perform well today is Language/Txt/Comments, Sound and Pic/Video.

    So, using AI for perusing social media and general stock/company sentiment is an obvious move. A very good move. This obviously includes Stock Forums of all types. However, I would excluded stocktwits and be careful of forums with massive bot useage and short comment bot attacks. It would be good if the AI could determine a bot or a general spam account and exclude it.

    Infact, AI can be used to analyze any company report, income statement, Guidance and PR statements that occur. But the social media sentiment is the finest use of its current level of development.

    Maybe the best use might be to feature trend changes within sentiment monitoring.

    You can also use AI for word searches on smaller less known companies, to quick search for certain hot sectors or hot industries.

    All of the above is how hot money moves today. Verbal sentiment, followed by vivid graphics and picture emotion. Injecting sound bits into the emotion is just around the corner. The new emotional traders and investors today have far too LOW of an IQ range to do any real analysis. And social media dumbs it down just perfectly.
     
    Last edited: Mar 17, 2025
    #21     Mar 17, 2025
  2. LisafHall

    LisafHall

    A lot of traders underestimate how important it is to fine-tune their models based on changing market conditions. Just throwing a general AI model at trading without constant adjustments is a recipe for disaster. I’ve had much better results after making sure my data sources stay relevant and adapting my approach when needed. Also, I’ve been testing different ways to integrate xonedium into my process to see if it can improve execution speed.
     
    #22     Apr 8, 2025
  3. It is refreshing to see someone openly explain both the strengths and the real-world challenges of integrating AI into trading. Many underestimate the complexity involved, especially when it comes to maintaining models, managing risk, and handling the operational side like APIs and infrastructure. Your point about AI helping more with decision support rather than pure prediction really resonates with my own experience. I also appreciate your honesty about the psychological hurdles — sometimes overcoming our own biases is the hardest part.