8 episodes
AI Sphere Ramasubramaniam
-
- Technology
AI Sphere is a podcast about the world of AI and associated technologies. The host is a business school professor and an avid researcher in the field of supply chain, logistics with special interest towards applications of machine learning, deep learning, and artificial intelligence. You can connect with the host in LinkedIn: https://bit.ly/3Gy3Gnb.
Disclaimer: The content in this podcast is purely the host's/co-host's opinions and is not representing any other individuals or organizations.
-
Understanding the Judges: Examining Indian Legal Judgements through Multilingual Summaries Dataset
In this episode, we review an arXiv paper titled "MILDSUM:
A NOVEL BENCHMARK DATASET FOR MULTILINGUAL SUMMARIZATION OF INDIAN LEGAL CASE
JUDGMENTS" by DEBTANU DATTA ,
SHUBHAM SONI , RAJDEEP MUKHERJEE and SAPTARSHI GHOSH". https://arxiv.org/pdf/2310.18600 -
Top AI Trends to Watchout in 2024
In this closing and shortest episode of AI Sphere for 2023, we take a sneak peek into potential AI trends to watchout in 2024.
-
AI and Google Layoffs in Advertising
In this episode, we delve into the recent news in the topic of AI and employment, the potential layoffs by google for its ad sales department. https://www.theinformation.com/articles/google-plans-ad-sales-restructuring-as-automation-booms
The statements made towards the end of this episode are purely opinion of the host. -
Cracking the Code: Can Project Q* Lead to AGI?
Welcome to today's episode of AI Sphere, where we'll be
exploring the recent controversial topic in the world of AI -The artificial
general intelligence, or AGI. AGI refers to a hypothetical AI system that
possesses the ability to understand, learn, and apply knowledge across a wide
range of tasks, much like human intelligence. It has sparked intense debates about its potential impact on job displacement and broadly on the society. -
GPT 3.5 Turbo 0613 vs GPT 3.5 Turbo 1106
In this episode, we compare the performance of ChatGPT turbo 0613 with ChatGPT turbo 1106. We experiment with API based RAG system, chatting with pdfs which address the area of machine learning fairness and present qualitative results of the comparison. Enjoy listening!
-
When and how to fine tune GPT 3.5 Turbo - Fine tuning 101
In this episode, we will explore the process of fine-tuning GPT 3.5, an advanced language model. We will introduce the concepts of zero-shot, few-shot, and multi-shot learning,
and discuss how they relate to fine-tuning. We will also cover the use cases for GPT 3.5 fine-tuning and the steps involved in the fine-tuning process. Fine-tuning allows developers to customize GPT 3.5 Turbo to suit different use cases, improving its performance on specialized tasks. With this podcast, you will gain a comprehensive understanding of the latest developments in AI and how to unlock its full potential.