Welcome to Future Work.

Every week, I scan the news for must-know stories about the employee-centric, happier, distributed, and AI-driven future of work.

Below is this week’s edition. Not a member yet? Join over 10,000 people-centric managers and subscribe here.

Read listen? The spoken version is on YouTube, Spotify, and Apple Podcasts.

Have you created your own "GPT" yet?

In Future Work #61, that's what we'll explore.

The Need-to-Know:

• Customizable AI Capabilities: OpenAI has introduced GPTs, customizable versions of ChatGPT. You can train them to perform tasks aligned with individual, team, or company needs.
• Save Your Time: These AIs can streamline workflows, increase efficiency, and offer direct interaction for feedback and problem-solving, reducing manual effort and time in any process.
• Practical Implementation: Initial GPTs developed by OpenAI let you explain Gen Z memes and negotiate contracts. Community applications include GPTs in editing, HR, interviewing, management, and personal development.
• It's Easy: Building a personalized GPT is accessible to those with ChatGPT Plus or Enterprise and is simpler than ever imagined.
• Future Potential: The potential of these GPTs is significant, as they represent a move towards AI agents capable of handling multiple steps in work processes, indicating a future where AI could increasingly take over structured, repeatable tasks.
  ‍

But wait... Did you still not sign up for Reclaim.AI? (Sponsored)

Reclaim is an AI scheduling automation app that finds the best time for your meetings, tasks, habits, & breaks.

Square, GitHub, Zendesk, and 30,000 other companies use their Smart AI scheduling technology for busy teams.

Try it today, for free.  

PS: FlexOS is opening up its platform for a first month-long sponsor takeover. Want to reach 20k+ tech-forward, people-centric leaders? Contact us.
‍

Smart Managers Create Their Own GPT Today (With 50+ Examples)

Overnight, the world has changed.

AI can now actually do all your work for you. 

(Note: AI is developing rapidly. Take a look at our six must-know AI trends in 2024.)

Introducing: GPTs

What if you were to dream up your own AI to do exactly what you want it to do? Or simply thinking how to use AI?

As announced at OpenAI’s demo day, that is now possible.

And they’re called GPTs. (As an aside, what a terrible name.)

So, what are GPTs?

GPTs are custom versions of ChatGPT, purpose-trained for your particular objective, fueled by up to thousands of pages of knowledge you have on any topic or skill.

OpenAI created a few examples to use and inspire you:

• Game Time: can quickly explain board games or card games to players of any age. Let the games begin!
• The Negotiator: helps you advocate for yourself and get better outcomes. Become a great negotiator.
• Tech Support Advisor: from setting up a printer to troubleshooting a device, they’re here to help you step-by-step. (Send this to your parents after explaining ChatGPT to them :)) 
• Math Mentor: helps parents help their kids with math. Need a 9 pm refresher on geometry proofs? I’m here for you.
• genz 4 meme: helps “u understand the lingo & the latest memes” - critical in case you don’t speak Gen Z

The simplest way to think about it is that you can train it on anything you would otherwise have done yourself manually.

For example, if you usually review the work of direct reports based on your knowledge, upload that knowledge to a GPT.

Rather than checking the work manually, you can let the GPT review it for you.

You can even let your direct report interact with your GPT to get feedback and fix issues before submitting it to you.

As Kyle Wiggers writes in Techcrunch, the power of these GPTs is incredible beyond expectations:

“The startup telegraphed its ambitions in March with the launch of plugins for ChatGPT, its AI-powered chatbot, which brought third parties into OpenAI’s model ecosystem for the first time. But what caught this writer off guard was the breadth and depth of OpenAI’s GPT building — and commercializing — tools out of the gate.” – Kyle Wiggers, TechCrunch

But the truth is, until you try, it’s hard to grasp exactly what this is all about. Iliya Valchanov, who taught AI to over 1 million users, explains it as follows:

“All the smart people say 'adopt GPT-4' but even they don't know WHY. The reason is that you need to speak to ChatGPT and GPT-4 in particular to understand it. No one can explain it to you. It is a feeling (yes, like love but not love). It's a feeling like.. holding a tennis racket, or a fork, or driving the car. You just 'feel' it. There is a barrier that we need to overcome before we become one with the car GPT.” – Iliya Valchanov, CEO, Team-GPT

A practical example of how GPTs deliver value: the AI Editor

Let me give you an example of a workflow we have significantly streamlined since GPTs launched last week:

To publish articles on our website, we go through a rigorous editing process to publish articles.

A significant part of the editing process is reviewing the submitted work of a writer against a checklist of requirements.

These requirements include the tone of voice, sentence lengths, and which research references we accept.

Checking articles against our checklist would typically be a manual process of submitting, reviewing, debriefing, revising, etc.

Now, a GPT can do this work for us.

Even better, the writer can directly interact with the "GPT Editor" and self-service the editing process before submitting to us.

Shortcutting this process creates a ton of value because every step of the editing process adds time between writing and publishing.

These time savings mean we can create more articles or spend less money on our content. It also means the editor can focus on higher-value tasks.

A real game-changer.
‍

Build Your Own

Building your own GPT is not difficult.

I repeat, it’s NOT difficult. 

All you need is ChatGPT Plus or Enterprise, and you’re off to the races. 

Here’s a nice tutorial on how to do it:

Let me add a step-by-step example below.

I created a GPT for writing your Personal User Manual – a great tool to improve collaboration. 

Here's how I did it. 
‍

Step 1: Define the Purpose 

The GPT Builder asked me what this bot is supposed to do. You can see in the screenshot that it immediately ‘gets’ what I’m referring to –AI magic!

Step 2: Train It 

After setting a name and a profile image, we get to work on training the bot. 

I explained what I wanted the bot to do and uploaded relevant documents that should fuel the ‘knowledge’ of the bot. This ensures that GPT follows a specific framework and source of knowledge versus everything it knows. 

Providing this data makes it personalized.

GPT reads all the data you upload. In this case, a 2000-word PDF, but it could also be 5 books you’ve written. 

It’s how you can have an AI trained on only your or your company’s data. 

Step 3: Test It!

Shockingly, that’s it!

That’s all it took to create a customized ChatGPT-style AI based on your specific knowledge, processes, or whatever else you’d like to create a bit for.

OpenAI lets you test your new GPT in the right frame to see how the GPT could be improved. Instead of editing your bot manually, you can speak to it, and it will improve itself.

For example, my GPT asked many questions all at once, which could be more effective in crafting a Personal User Manual with the user.

So, I told GPT to ask the questions one by one in a conversational style. 

Within three seconds, the user experience of the bot increased by 1000%. This illustrates these GPTs' power – and this is only their first iteration.

Interested to try it out? The “Personal User Manual Bot” is now live here.

But more importantly – get started to make your own. Creating AI Agents will be a key future skill. At least while we still have jobs. 

For more inspiration, Hernan Chiosso created a GPT Creator Canvas: 

Important Disclaimer

Please note that this is an early product release and that issues may occur. 

Specifically, experts like Jacques Alomo highlight not uploading confidential information and data in either the instructions or the attached files, as users may be able to retrieve them. 

Additionally, it’s good to understand that this will feed OpenAI with more training data without needing to pay for it from big publishers. The adage still holds: “If you don’t pay for the product, you are the product.”

Great Examples of GPTs to try Today

For Managers

• Thomas To from our team created a GPT to get advice on any management challenges based on our database of best practices: https://chat.openai.com/g/g-UT6K00O0B-lexi-coach 
• Michael Preuss created a strategic advice GPT with “the wisdom of top-tier MBAs and elite consulting firms”: https://chat.openai.com/g/g-4xw63Av3D-action-planner 
• Stephanie Lee created a(n assistant to the) Head of Remote: https://chat.openai.com/g/g-tyXsOQcuT-assistant-to-the-head-of-remote 

For HR

• Lars Schmidt created a GPT to advise you on how to GPT your HR processes (very meta indeed): https://chat.openai.com/g/g-0sMUzPz7U-hr-automation-gpt 
• Lars also created a GPT that allows you to get feedback on and improve your HR strategy: https://chat.openai.com/g/g-UWD9NXKHN-hr-business-strategy-copilot 

For Anyone 

• Decision-Maker Bot: David Arnoux created an assistant for making informed and strategic decisions in your personal and professional life: https://chat.openai.com/g/g-P9aIJx2VR-decision-maker-bot
• Design: Jason Culbertson of Cameo created a “What Would Apple Do” GPT, where you can get design advice based on Apple design guidelines: https://chat.openai.com/g/g-qwp9jR7Fd-what-would-apple-do
• Creating Websites: Pietro Schirano created a GPT that turns instructions into full, hosted websites: https://chat.openai.com/g/g-2Eo3NxuS7-designergpt
• Career Advice: Chris Grubbs of KGG created a career coach. I gave it a spin, and it created a helpful plan: https://chat.openai.com/g/g-qgiJ7GjGP-career-coach 
• Interview Coach: Yogesh Chavda created a GPT to help job seekers prepare via mock interviews. https://chat.openai.com/g/g-X51kqvpxq
• Better Prompting: AI coach Mark Fulton created a GPT that helps you improve your ChatGPT prompts: https://chat.openai.com/g/g-UiIlbfThK-smartgpt 

For fun

• Job from Remote created a coloring book GPT. You can even feed it reference photos, like my son’s Halloween outfit: https://chat.openai.com/g/g-GyuHDtXnl-coloring-crafter

• Ammaar Reshi at Brex created a movie recommendation GPT: https://chat.openai.com/g/g-Gm9cCA5qg-what-should-i-watch 
• Matt Schlicht from Octane lets you create a Simpsons character from a photo: https://chat.openai.com/g/g-tcmMldCYy-simpsonize-me 
• Douglas Voeten from Perfect Dubs created a GPT to “roast” your photo. Apparently, I look like a car salesman: https://chat.openai.com/g/g-a14Q8Cybp-roast-master-supreme/
  

• Kevin Buehler created “Modern Shakespeare,” translating Shakespeare’s plays into modern, casual English: https://chat.openai.com/g/g-Ncy8liDLE-modern-shakespeare 

Others: 

• Futurist Antony Slumbers created a GPT that lets you negotiate leases: https://chat.openai.com/g/g-MQops8DHF-the-lease-negotiator 

You can find 50 additional GPTs on my LinkedIn post. I’m happy to continue expanding on this list – submit yours here.

The Bottom Line

Any work you do that’s repeatable and structured, an AI will take over. You can either wait for that to happen mainstream – or get ahead today. 

Creating GPTs is an easy way to streamline your workflows and increase opportunities for your teams to get more of your support without additional management time.

As Ethan Mollick writes, GPTs are a first step towards agents – GPTs that can go through multiple steps of a work process and truly take over your work.

Smart managers take advantage ahead of the rest. 

And I know you’re one of them.

‍

– Daan

‍

Previous articles on AI at Work:

• A Jobless Future? Interrogating Musk's AI Prophecy 
• ChatGPT can now See, Hear, and Speak – How to Stay Ahead in the Age of AI?
• Remote Work, AI, and Productivity
• Three reasons why most of us still don't use A.I. (and keep wasting our time and energy.)
• From Coders to Writers: Jobs AI Will Replace
• Communication Overload is crushing us. Can A.I. fix it?
• How AI Will Transform Management (+ 27 Tools)