Ultimate AI Masterclass for Founders and Executives 💰| Vaibhav Sisinty | IBP EP28
Executive Summary
Think School host Ganesh Prasad talks with Vaibhav Sisinty, founder of Growth School, about how working professionals and founders should use AI agents as layoffs and a possible US recession put pressure on jobs. Sisinty explains AI agents versus chat assistants with an overnight task using ChatGPT Operator, why he thinks companies will replace costly layers of management with AI processes, and a leveling roadmap: define your toolkit, learn how models and prompts work, and apply them to real tasks. Coverage comes from approximately the first 50,000 characters of third-party captions of a 1 hour 41 minute episode from April 2025, so later parts are not covered. Model names and predictions (for example code written by AI, recession timing, layoff counts) are the speakers' statements from 2025, are partly dated, and were not independently checked. This is not career, financial or legal advice. Captions are automatic with errors. Topics are untimed; metadata follows YouTube. Think School lists Odoo as a sponsor. Source: https://youtubetotranscript.com/transcript?v=Rt7qxyDzc_Y
Key Takeaways
The setup: layoffs and recession fear
The host says he fears a US recession and notes layoffs at Amazon and Google. He asks what a working professional should do. Sisinty says weaker consumer spending leads companies to cut people, and a recession gives them a reason to adopt efficient AI-based processes.
Agents are not assistants
Sisinty says people think of AI as a chat assistant, but the real power is that it can complete a full task on your behalf. He describes using ChatGPT Operator overnight to go through about 5,000 Y Combinator companies and return 25 to 26 ideas, needing him only to log into Gmail.
Why he thinks middle managers are at risk
He argues companies have a window to replace expensive humans with AI agents and are cutting middle and senior managers first. This is his view, offered as a prediction.
Code and models
He says 60 to 70 percent of the code in his company is written by AI, cites a Y Combinator survey with some companies at 90 percent, and says people are using older models. These are his claims from 2025 and the model names are now dated.
Level zero: define your toolkit
He describes a leveling roadmap he used with his team of 20 to 25 over nine months. Level zero is defining your toolkit and checking tools through review sites like Product Hunt.
Prompting and research
He says a prompt like do research on US recession is the worst; give context about what you want to cover so the model has more surface area to think. He says deep research tools can explore more angles.
Know which model to use
He uses the example of preparing for a first investor conversation, saying you should know which model type fits the task, because a text-generation model differs from reasoning models. His point is that understanding how models work prevents misuse such as uploading a long PDF to the wrong model.
The Conversation Breakdown
The setup: layoffs and recession fear
⏱ Untimed topicThe host says he fears a US recession and notes layoffs at Amazon and Google. He asks what a working professional should do. Sisinty says weaker consumer spending leads companies to cut people, and a recession gives them a reason to adopt efficient AI-based processes.
Agents are not assistants
⏱ Untimed topicSisinty says people think of AI as a chat assistant, but the real power is that it can complete a full task on your behalf. He describes using ChatGPT Operator overnight to go through about 5,000 Y Combinator companies and return 25 to 26 ideas, needing him only to log into Gmail.
Why he thinks middle managers are at risk
⏱ Untimed topicHe argues companies have a window to replace expensive humans with AI agents and are cutting middle and senior managers first. This is his view, offered as a prediction.
Code and models
⏱ Untimed topicHe says 60 to 70 percent of the code in his company is written by AI, cites a Y Combinator survey with some companies at 90 percent, and says people are using older models. These are his claims from 2025 and the model names are now dated.
Level zero: define your toolkit
⏱ Untimed topicHe describes a leveling roadmap he used with his team of 20 to 25 over nine months. Level zero is defining your toolkit and checking tools through review sites like Product Hunt.
Prompting and research
⏱ Untimed topicHe says a prompt like do research on US recession is the worst; give context about what you want to cover so the model has more surface area to think. He says deep research tools can explore more angles.
Know which model to use
⏱ Untimed topicHe uses the example of preparing for a first investor conversation, saying you should know which model type fits the task, because a text-generation model differs from reasoning models. His point is that understanding how models work prevents misuse such as uploading a long PDF to the wrong model.
Core Frameworks & Ideas
Assistant versus agent - Sisinty's distinction
An assistant answers a question; an agent completes a multi-step task on your behalf.
Level zero to higher levels - Sisinty's roadmap
Start with your toolkit, then learn how models and prompts work, then apply them to real work.
Surface area in prompts - Sisinty's tip
Give the model your context, goal and angles so it can explore more than a thin one-line request.
Actionable Next Steps
- ✓Hand an AI agent one multi-step task this week and check what it gets wrong.
- ✓List the tools you use and look up reviews for each before adopting new ones.
- ✓Rewrite one of your prompts to include your goal, audience and angles.
- ✓Check which model type fits the task before you start.
Notable Quotes
“the real power of AI today as we live is that it has the ability to complete a full task on your behalf”
“you have to start leveling yourself up as an AI generalist”
“you're giving it more surface area to think and give you a better output”
The One Thing to Remember
AI agents can complete whole tasks, so learn them as a generalist before your employer decides what your role is worth.