Week 5-A Supervised Approach To Training An Artificial Intelligence To Extract Relevant Genomic Data From Literature

Adam B -

Hello,

This past week, I’ve been working on finalizing the API that will automate the analysis my AI model uses. This is essential to address the sheer size of psychiatric literature within a reasonable span of time.

Major Developments

The API has improved dramatically from when I first began last week. It can now process a variety of human-based inputs, allowing analysis to be targeted toward specific sectors of psychiatry. This enables the possibility of both: a) uncovering unlikely relationships between topics that otherwise would’ve been too costly or intensive to justify, and b) specializing information to eliminate any noise within findings. Additionally (and perhaps most critically to call anything autonomous), it can now extract multiple papers simultaneously without further user prompting. This drastically increases efficiency and scalability, moving the project from a theoretical proof-of-concept toward a truly functional research tool.

Learning Curve

This week, I worked more with Python. Much of my work involved dissecting code to ensure I truly understood how my program works. The rest was spent coding with AI, leveraging its capabilities to troubleshoot bugs and quickly output viable code. One of the biggest lessons I learned is that AI-assisted programming is most effective when you understand the fundamental logic of the system you’re working with.

What’s Next?

With the API nearly finished, the next major step is validation: The API will be run on a controlled set of psychiatric papers, and its outputs will be manually verified against human-extracted data. The difference between this and the training phase mentioned in prior weeks is that validation should not require rewriting of the prompt and will be done in bulk (as opposed to the one paper -> revise process during training). If successful, this will provide the first concrete proof that my AI model can reliably assist in psychiatric research.

Thanks for following along,

Adam

More Posts

Comments:

All viewpoints are welcome but profane, threatening, disrespectful, or harassing comments will not be tolerated and are subject to moderation up to, and including, full deletion.

    camille_bennett
    Hi Adam, really cool to see how much the API is improving. I know that you plan to study medicine long-term. Do you feel your exposure to AI through this project will help you on that path?
    March 10, 2025 at 10:17 am - Reply
    adam_b
    Hi Ms. Bennett. Yes! I think AI has places in a variety of different fields, medicine included, so learning how it's being incorporated in its early stages is very good experience.
    March 11, 2025 at 2:26 pm - Reply
    tanay_n
    Hey Adam, this is really exciting! Are there any ethical considerations regarding AI-driven psychiatric research that you’ve had to address?
    March 13, 2025 at 3:52 pm - Reply
    adam_b
    Hi Tanay, that's a great question! Generally, it is important to have clear safety precautions and protocols when working with AI, especially in the medical field. This particular AI is processing information that is already publicly available and operates under clear human-overseen guidelines, so there are not too many ethical concerns with regards to patient confidentiality or data privacy.
    March 14, 2025 at 3:48 pm - Reply

Leave a Reply

Your email address will not be published. Required fields are marked *