30 January 2024

How To Replace Any Employee With Custom GPTs

1Frameworks

6Insights

Frameworks in this episode

Productivity6 steps

Action-Connected Custom GPT Agent

Turn a recurring software workflow into an callable AI specialist.

Read →

Insights & moments

The myth-busts, hot takes, explainers, and tools worth keeping.

Myth Buster· 1

Myth Buster02:30

A Custom Prompt Does Not Automatically Make a Valuable GPT

Many custom GPTs simply package generic instructions, personas, or quote-generation prompts around capabilities ChatGPT already has. The host says differentiation requires proprietary data or another capability that makes the GPT genuinely unique.

Pre-programmed instructions are easy to reproduce.
Personas alone rarely create defensible utility.
ChatGPT already has broad internet-trained knowledge.
Proprietary data can provide meaningful differentiation.

“if you do not have propriety data or something that makes that custom gbt unique then it's kind of underwhelming”
Host · 03:00

#custom gpts#prompting#differentiation#proprietary data

Hot Take· 2

Hot Take05:00

Automatic GPT Selection Could Undermine the GPT Store

The host identifies a tension between convenience and marketplace value. If ChatGPT automatically chooses the best custom GPT for every task, users may have little reason to browse featured or trending listings in the store.

Automatic routing could improve user convenience.
Invisible selection reduces deliberate store browsing.
Featured and trending placements depend on user discovery.
The interface and marketplace incentives may conflict.

“chat GPT would pick what custom GPT to to actually use and so what's the point in having this store”
Host · 05:00

#marketplaces#product strategy#gpt store#routing

Hot Take09:30

AI Agents May Become Callable Colleagues in Team Chat

The host predicts that future work interfaces will mix humans and specialized AI agents in the same conversational environment. Workers could mention either kind of participant and delegate an appropriate task without leaving the shared workspace.

AI agents may be addressed like coworkers.
Human and AI participants could share one interface.
Each agent may handle a specific type of work.
Conversational delegation could become a normal work pattern.

“you will probably have people you converse with uh who are humans in at them and people who are AI agents and at them”
Host · 10:00

“they can actually help us do our work”
Host · 09:30

#ai agents#future of work#slack#collaboration

Explainer· 2

Explainer01:30

Why Low Barriers Flood the GPT Store With Weak Tools

The host argues that creating and publishing a custom GPT is so easy that the store quickly accumulates tens of thousands of uneven offerings. Trending and featured sections help discovery, but they do not solve the underlying quality problem.

Publishing a custom GPT requires little effort.
Low barriers increase the volume of undifferentiated tools.
Discovery becomes difficult when useful GPTs are buried in noise.
Trending lists act as a partial quality filter.

“it's just too easy to create a custom GPT and launch it on the store”
Host · 01:30

“it's really hard to find like the good ones”
Host · 02:00

#gpt store#discovery#marketplaces#quality

Explainer03:30

Why Copying Between GPTs Breaks the Workflow

The host demonstrates how a user might generate a hopeful quote and then want to reuse it in a Canva presentation. Separate GPT interfaces force the user to switch tools and manually transfer material, making a potentially useful workflow cumbersome.

Different GPTs may contribute to one final deliverable.
Separate interfaces fragment conversational context.
Manual copying adds friction and error opportunities.
A central interface makes specialist tools more usable.

“I has to kind of flick between them and it's really really annoying right”
Host · 04:30

#context switching#user experience#canva#workflow

Takeaway· 1

Takeaway11:30

Why GPT Mentions Could Be 2024's Sleeper AI Feature

The host believes invoking custom GPTs from the core ChatGPT interface will make them practical for daily use and increase the value of the GPT Store. He expects users to customize ChatGPT around their needs by assembling multiple specialized agents.

In-chat invocation reduces interface friction.
Daily use matters more than novelty.
Specialized agents can customize a general AI workspace.
The feature may increase the practical value of the GPT Store.

“this is a game changer for how we feel about custom chpts in the GPT store”
Host · 12:00

“we're going to have an army of AI agents to be able to do work on our behalf”
Host · 12:30

#chatgpt#custom gpts#productivity#ai trends