Recent talks held at two leading universities, titled “Knowledge in the Age of AI” and “Innovation for Impact” fueled a deeper interest in understanding GPT technology - its inner workings, applications, and potential. The discussion explored AI innovations - from how they are currently applied, where they are most needed, and how they are being developed. Here are the general notes on this.
Problem: Handwritten Digit Classification
— Trueblogposts (@trueblogposts) July 29, 2026
Goal: Build a machine learning model that can classify images of handwritten digits (0-9) based on pixel data.
Knowledge in the Age of AI
Seminar talk with Informatics Departmental
Q. How will knowledge representations fit into a world transformed by LLMs?
The deployment of large language models (LLMs) has changed the way we approach and think about knowledge representation. Traditionally, knowledge representation (KR) has focused on formalizing information in ways that machines can process and reason about, such as through ontologies, semantic networks, logic-based systems or knowledge graphs. With the rise of LLMs, we are seeing a shift in how knowledge is represented, organized, and interacted with.
To process language, the brain relies on neural networks, just like how a machine learning model uses artificial neural networks (ANNs) to understand and generate language. Both systems aim to recognize patterns in language (words, phrases, sounds) and predict what comes next based on learned experiences or data.
As people construct sentences, the brain analyzes what words or phrases are most likely to follow given the structure of the sentence and previous experiences with language. For example - The brain predicts how to complete a thought or sentence. In a conversation, if someone says, "I love to play soccer because it’s," the brain might predict they will say "fun" or "exciting" based on the context of the conversation.
Seminar talk with Informatics Departmental
In recent years, large language models have greatly impacted research and products across various fields, from healthcare to soda flavor development. While early excitement may have been overstated, generative AI is undeniably a transformative technology with a profound impact, though its full extent is still unclear.
Q. What role do knowledge infrastructures, graphs and metadata repositories play in this new world?
In this world of rapidly advancing technology, interconnected systems, and growing data volumes, knowledge infrastructure, metadata repositories, and knowledge graphs play a big role in enabling data management and decision making.
A knowledge graph for example is a structured representation of knowledge that captures relationships between different concepts or entities. They represent entities (like people, places, products, etc.) as nodes, and relationships between them as edges. Knowledge graphs use semantic technologies to organize data and uncover deeper insights through pattern recognition and relationship mapping. Knowledge infrastructures on the other hand provide the foundation systems for knowledge creation, sharing and utilization across industries.
Knowledge graphs can integrate clinical, genomic, and experimental data to help identify patterns, relationships, and insights that may not be apparent in isolated datasets. Metadata repositories be utilized to manage data from various sources. This includes brain imaging, genomics, patient records, and clinical trials.
Q. How will knowledge representations fit into a world transformed by LLMs?
The deployment of large language models (LLMs) has changed the way we approach and think about knowledge representation. Traditionally, knowledge representation (KR) has focused on formalizing information in ways that machines can process and reason about, such as through ontologies, semantic networks, logic-based systems or knowledge graphs. With the rise of LLMs, we are seeing a shift in how knowledge is represented, organized, and interacted with.
To process language, the brain relies on neural networks, just like how a machine learning model uses artificial neural networks (ANNs) to understand and generate language. Both systems aim to recognize patterns in language (words, phrases, sounds) and predict what comes next based on learned experiences or data.
As people construct sentences, the brain analyzes what words or phrases are most likely to follow given the structure of the sentence and previous experiences with language. For example - The brain predicts how to complete a thought or sentence. In a conversation, if someone says, "I love to play soccer because it’s," the brain might predict they will say "fun" or "exciting" based on the context of the conversation.
Innovation for Impact
Symposiums with BioScience Futures
Symposiums with BioScience Futures
Medical innovations, cloud infrastructure, and artificial intelligence (AI) are reshaping healthcare at a fast pace. The technologies work in connection together to drive advancements in medical research, patient care, and healthcare management.
Q. What role does machine learning play in Life Sciences and structure prediction?
Machine learning is being applied in areas such as protein folding, RNA secondary structure prediction, and drug design. One of the most notable successes of ML in biological structure prediction is AlphaFold, developed by DeepMInd. The goal is to predict the 3D structure of a protein based on its amino acid sequence. Predicting protein structure has been a major challenge in biology because the number of possible configurations of a protein is astronomically large.
AlphaFold uses deep learning techniques to model the physical and chemical properties that govern protein folding. By training on known protein structures, AlphaFold can predict highly accurate 3D structures for proteins that have not been experimentally solved. This has led to breakthroughs in understanding diseases, drug design and biological pathways! In 2021, AlphaFold was able to predict the structure of all proteins in the human genome with great accuracy.
Cloud infrastructure includes physical and virtual resources (servers, storage, networking, etc.) that are hosted on remote servers (data centers) and accessed over the internet (the cloud). This includes the storage, processing, and management of vast amounts of data like patient records, diagnostic images, and treatment histories.
The integration of Cloud computing and AI in healthcare is currently in effect as Telemedicine and Remote Consultations. Cloud-based healthcare platforms enabling remote consultations, reducing the need for patients to travel long distances to see specialists. Cloud computing enables healthcare professionals to collaborate across geographic boundaries, sharing knowledge, research, and medical data seamlessly. Ai-powered tools are currently assisting doctors during these remote consultations.
“Mere models, understanding systems”- Chatbots are not actually understanding the content they generate, but simply following patterns learned during training.
— Trueblogposts (@trueblogposts) March 20, 2025
Additional Topic Research
Face-to-face and digital articles
Neurotech is a rapidly growing field. It focuses on developing technologies that interact with the nervous system, aiming to enhance or restore brain function, monitor brain activity, and address neurological disorders. Q. Examples of bridging the gap between Mind and Medicine?
Psychosomatic medicine examines how psychological factors impact physical health, with conditions like chronic pain, gastrointestinal issues, and heart disease often worsened by anxiety or depression. Doctors in this field address both mental and physical health. Psychoneuroimmunology in contrast, studies the connection between the brain, immune system, and mental health, showing how stress can affect immune function and contribute to illness. This has led to treatments like stress reduction techniques and therapy.
Natural Language Processing (NLP), wearable sensors, big data and analytics, and Deep Learning are key technologies that will support these applications. NLP can be applied for analyzing text-based data like patient records, therapy notes, or journal entries to understand psychological states, to Deep learning for processing complex data, such as imaging (e.g. brain scans) or time-series data (e.g. heart rate).
Q. What's the discourse on advanced tools and techniques?
Brain-Computer Interfaces (BCIs) lead into the broader field of Neurotechnology, which includes neurostimulation, neuroprosthetics, neurodiagnostics, and monitoring.
These technologies can help reduce physical strain in various ways. For example, BCIs may minimize the need for repetitive motions, potentially preventing issues like carpal tunnel syndrome or repetitive stress injuries. For workers with disabilities, neuroprosthetics offer the chance to regain lost functionality, making workplaces more accessible.
For individuals dealing with mental health conditions, whether short-term or long-term, Deep Brain Stimulation (DBS) or neurostimulation could offer therapeutic benefits, enhancing emotional well-being and promoting a better work-life balance, potentially reducing burnout. Additionally, neuropharmaceuticals like nootropics could help with improving memory, learning, and decision-making skills, ultimately boosting their performance.
So.. Does it makes sense?
Many machine learning algorithms are inspired by the way the brain works. For example, deep learning models are based on the brain's layered structure for solving problems and processing information hierarchically. The field of neuroscience has also played a key role in driving the development of more interpretable AI. How? Techniques like attention mechanisms and neural symbolic networks in the past have now helped us understand how AI models make decision. This all links back to our understanding of how the brain makes its own decisions!
Resources
(Article) Empowering people to unlock AI’s full potentialNearly all companies invest in AI, but only 1 percent feel they have reached full maturity. The firm's research shows that the main obstacle to scaling is not the employees who are ready—but the leaders, who are not moving quickly enough.
In the ever-evolving landscape of scientific research, the intersection of neuroscience and artificial intelligence (AI) highlights the strength of interdisciplinary collaboration.
Discover the latest research in Neuroscience, including insights on how brain activity can enhance workplace culture, the application of neuroscience in leadership training, and the connection between organizational trust and business performance.
Last updated: 01/01/2026
.png "Baby laughing while reading a book")
