
Sign up to save your podcasts
Or


What happens when AI can do the work of a junior analyst? And what happens to businesses when software becomes dramatically cheaper and faster to build?
In this episode, Paaras speaks with Shantanu Shah, founder of empor.top and HeyFinn.co, about how AI is transforming two worlds: building businesses and investing in them.
Shantanu is building tools that help investors research companies, analyze qualitative information, and automate parts of the investment research process. Drawing on his experience building Bewgle, empor.top, and HeyFinn, we explore both the possibilities and the limitations of AI.
Whether you're an investor, entrepreneur, or simply curious about the practical implications of AI, I hope you find the conversation interesting.
Chapters
00:00:00 Introduction
00:02:04 The creation of empor.top
00:10:04 HeyFinn: An AI junior analyst
00:15:05 Qualitative research and lessons from Bewgle
00:23:32 How AI changes team productivity
00:29:20 Can a one-person company be worth $1 billion?
00:36:10 Humans as curators and inspectors
00:40:45 What LLM intelligence cannot solve
00:49:12 Multiple models and personalized analysis
00:56:03 Frontier models and thematic research
01:01:17 Why AI software still takes time to build
01:14:01 Where will software moats survive?
01:19:29 Specialist AI versus foundation models
01:22:27 Process-driven software moats
01:30:05 How investors should use AI
01:37:01 What machines do better than analysts
01:42:31 Getting investors to delegate to AI
01:47:31 Hallucinations, feedback, and trust
01:51:30 Primary sources and independent conviction
01:54:41 Can AI think against the consensus?
02:02:48 Complementing an analyst's strengths
02:06:23 What differentiates human investors?
02:10:46 Automating repetitive intellectual work
02:15:32 HeyFinn's open questions and the name Empor
02:21:39 Lateral thinking and patience
Explore Shantanu's work:
empor.top — https://empor.top
HeyFinn — https://heyfinn.co
If you enjoyed the conversation, consider sharing it with someone interested in AI, investing, or building businesses.
Summary
In this conversation, Rish, a software engineer with extensive experience, shares his journey from building apps in university to working at major tech companies like Google and Revolut.
He discusses the evolution of AI in engineering, its potential to enhance productivity, and the challenges it presents. Rish emphasizes the importance of understanding the fundamental problem being solved, highlighting how AI tools can streamline operational tasks while also introducing new complexities.
In this conversation, the speakers delve into the intricacies of using AI tools in programming, the evolution of AI models, their impact on development timelines, and the economics of token usage. The conversation also touches on the trustworthiness of AI in various tasks, and emphasizes the importance of human oversight in the development process.
Chapters
00:53 Introduction to AI in Engineering
13:57 Professional Experience
22:04 Recognizing the Potential of AI
34:35 The Good and Bad of AI in Development
39:44 Understanding AI harnesses
47:26 Comparing AI Tools: Codex vs. Claude vs. Gemini
53:59 The Impact of AI on Development Timelines
01:01:09 Token Economics and AI Usage
01:07:52 Personality and Performance of AI Models
01:15:43 Trusting AI: Limitations and Concerns
01:24:47 The Challenges of AI in UI Design
01:30:03 Human Judgment vs. AI: Where Do We Draw the Line?
01:34:47 The Impact of AI on Managerial Expectations
01:52:40 The Evolution of Software Engineering Products
02:01:17 The Future of AI in Software Engineering
02:12:25 Future of Engineering and Skills Evolution
02:21:22 Citizen Developers and Problem Solving
02:27:38 Investor Perspectives on AI and Software