Skip to content
5Mins.ai Header
AI Declassified - Episode 2

The Jurassic Park Problem

How Not to Use AI - 7 Mistakes Even Daily Users Make

"Your scientists were so preoccupied with whether they could, they didn't stop to think if they should." Just because AI can do something, doesn't mean it should.

After two years of daily use, here are the 7 mistakes even daily AI users make - and the small fixes that 10x your output.

  • The settings 90% of users have never opened - and what to switch off today
  • Why "try again" is the worst thing you can type when AI gets it wrong
  • Browser vs. desktop app - the difference between the demo and the cockpit
  • Walk away with a one-page Power-User Cheat Sheet you can pin next to your monitor
Friday 15 May 2026 Available on demand
Saurav Chopra James Francis
Hosted by Saurav Chopra and James Francis
The Jurassic Park Problem: How Not to Use AI

What You'll Learn

7 mistakes even daily AI users make - and the small fixes that 10x your output

Everyone's running webinars on how to use AI. Almost no one is showing you the bad habits you've already built - the ones that make your AI feel mediocre when it should feel like magic. Saurav and James use these tools every day to run a B2B SaaS company. In 45 minutes, they'll show you the 7 things they've learned not to do - each one demoed live on screen-share, each one fixable in under a minute.

First, the honest bit

You're probably using AI like it's Google.

73% of knowledge workers use AI weekly. Almost none have changed their default settings, opened a project, or written more than a one-line prompt. One-shot prompts. Same chat for every topic. Default settings. Whatever the first reply was - that's what you ship. It's not your fault. Nobody trained you on this. This session is the second lesson you never got - the seven habits that separate someone who's used AI for six months from someone who's actually fluent.

Stop doing these. Start doing the opposite.
Mistake 1 Stop doing this

You never opened the settings panel

Memory's on. Training's on. Last Tuesday's LinkedIn rant is now context for this Tuesday's board deck. Most consumer AI tools default to remembering things across chats - and quietly using your conversations to improve the model.

The fix: Open settings. Look at memory. Delete the noise. Turn off training-on-your-data if you're on a consumer plan. Two minutes, once. Done.

Do this in the first 5 minutes of the session
Mistake 2 Stop doing this

You repeat instead of restart

Bad answer → "try again" → "be more specific" → answer gets worse. You're not correcting the model, you're poisoning the context. Every reply now anchors on the wrong direction.

The fix: When AI veers off, don't argue with it. New chat. Better prompt. Move on. The pros restart. The amateurs negotiate.

Mistake 3 Stop doing this

You're using the browser, not the desktop app

The browser version is the demo. The desktop app is the cockpit. It can see your screen, read your folders, and connect to Slack, Drive, HubSpot, and your calendar in real time. Almost no one knows.

The fix: Download the desktop app. Connect one tool today - start with Drive or Gmail. Live demo in the session.

Mistake 4 Stop doing this

You start from zero every conversation

"I'm the CEO of a B2B SaaS company in HR tech…" - and you type it again tomorrow. 80% of every prompt is the same boilerplate.

The fix: Build it once. Projects, custom instructions, skills, custom GPTs all exist for this. Load context once, reuse it forever.

Mistake 5 Stop doing this

You picked one AI tool and stuck with it

"Which AI is best?" is the wrong question. The right question is which combo. Claude for writing and reasoning. ChatGPT for image and quick answers. Perplexity for research with sources. Each has a sweet spot.

The fix: Don't subscribe to four. Pick a primary, then learn when to bounce a task to a different one.

Mistake 6 Stop doing this

You're trying to be a prompt engineer

Stop sweating over the perfect prompt. Ask the AI to write the prompt: "Help me write a prompt that will get you to do X really well. Ask me clarifying questions first."

The fix: Meta-prompting beats prompting. Two minutes of Q&A up front saves an hour of bad output downstream.

Mistake 7 Stop doing this

You shipped the first draft

The first output is the floor, not the ceiling. The second draft - where you go back and revise the prompt, not just the output - is where the real quality lives. "Now do that again, but cut the corporate tone, drop the bullet points, and add a specific example from financial services."

The fix: Treat output as a diagnostic. What went wrong tells you what was missing from the prompt. Add it. Re-run. Now it's good.

Your takeaway The cheat sheet

The Power-User Cheat Sheet

Every registrant gets a one-page printable PDF you can pin next to your monitor. Seven mistakes, seven fixes, the exact settings to change, and the bonus tips we don't have time to cover live. Bring it to your team's next AI working session. Beats another deck.

Bonus tips we'll squeeze in if there's time

  • Voice mode for thinking - walk-and-talk for 10 mins beats 30 mins of typing
  • Paste your calendar, not just your question - context beats cleverness
  • End every prompt with "what am I missing?"
  • Use temporary chat for sensitive stuff (the redundancy list, the offer letter)
  • The "explain it like I'm the auditor" stress test
  • The one keyboard shortcut that cuts your AI workflow in half
10×

Average user vs. fluent user. Same tool, same model, same monthly fee - wildly different output. The gap isn't the AI. It's the seven habits we'll fix in 45 minutes.

Don't learn the hard way. Watch the full session above and get the Power-User Cheat Sheet PDF.

Your Hosts

Meet the speakers

This session is hosted by two experts who use AI every day. They know where AI helps people, but more importantly where it should stay out of the way.

Saurav Chopra

Saurav Chopra

CEO & Founder, 5Mins.ai

Saurav is a serial HR tech entrepreneur on a mission to solve workplace challenges with technology. He previously co-founded Perkbox, the global employee experience platform now serving 5,000+ employers and 3M+ employees, before exiting to Great Hill Partners. His latest venture, 5Mins.ai - the "TikTok of workplace learning" - combines generative AI, microlearning, gamification and peer learning to deliver a highly engaging learning experience for HR and compliance teams. Saurav holds an MBA from London Business School and an engineering degree from IIT Delhi, and is the recipient of the Barclays Scale Up Entrepreneur of the Year award.

Connect on LinkedIn
James Francis

James Francis

Head of Growth Marketing, 5Mins

James is an experienced tech marketer with a background covering HR, e-commerce, property, and financial services. In his role as Head of Growth Marketing at 5Mins.ai, he leverages AI to deliver high-impact campaigns and accelerate growth.

Connect on LinkedIn

See how 5Mins.ai can transform your training strategy

Explore 5Mins further with our team and see why teams in 80+ countries love using 5Mins.ai for their training needs.

G2 Awards 2026 - High Performer, Users Love Us, High Performer Mid-Market
Footer Social Icons
Download_on_the_App_Store_Badge
Google_Play_Store_badge_EN

5Mins AI LTD
Ludgate House 107 - 111 Fleet Street 
London, EC4A 2AB, United Kingdom

 

© 5Mins AI Ltd 2026. All rights reserved

Thousands of teams trust 5Mins.ai with their compliance

EverBridgeLogo1_White-01
Kennedys Logo_Negative (White)_transparent background
gb-logo-white-mag
Dropbox-Logo
perkbox-logo-white-1
glaxosmithkline-white
london-business-school-white (1)
holidayextras white
st-jamess-place-logo-white