Prompt engineering

YouTube

Prompt engineering best practices are becoming obsolete

Dylan Davis March 20, 2025

---

Prompt engineering 10

AI tools & automation 7

• Reasoning models make chain-of-thought prompting redundant by default.
• Agentic tools generate their own steps, making explicit step prompts useless.
• Describe what you want — AI knows how to get there better than you do.

Read more →

AI tools & automation

YouTube

How to prompt reasoning models for better outputs

Dylan Davis January 28, 2025

---

AI tools & automation 10

Prompt engineering 9

• Your ChatGPT habits actively hurt performance in reasoning models.
• A six-section prompt structure unlocks far higher output quality.
• Embed a writing diff in the system prompt to clone your style.

Read more →

AI strategy & adoption

YouTube

Practical principles for using AI at work, with Ethan Mollick

Coaching for Leaders July 22, 2024

---

AI strategy & adoption 9

Prompt engineering 7

• AI's jagged frontier: great at sonnets, bad at exact word counts.
• Best prompters are managers, not coders — converse, don't code.
• Assume today's AI is the worst you'll ever use; urgency is real.

Read more →

AI tools & automation

Podcast

What's True and False About ChatGPT: A Practical Guide

Beyond the To-Do List - Productivity for Work and Life March 27, 2023

---

AI tools & automation 9

Prompt engineering 8

Automation & tools 6

• ChatGPT is probability-based software, not sentient AI—it excels when given detailed prompts but halluccinates beyond its knowledge.
• Prompt engineering (specifying role, task, and requirements) transforms mediocre outputs into enterprise-quality content.
• AI replaces tasks not jobs; entry-level roles face elimination unless companies redeploy freed time toward human-centered work.

Read more →

AI strategy & adoption

YouTube

Seven AI SaaS business ideas using the GPT-3 API

TK Kader January 22, 2023

---

AI strategy & adoption 9

Business models 7

Prompt engineering 6

• The moat is your prompts and training data, not the model itself.
• Fine-tuning GPT-3 on proprietary data compounds your competitive advantage over time.
• Seven concrete SaaS categories — from email automation to content personalisation — are buildable today.

Read more →