Je teste les GPTs sur COPILOT ! (mieux que sur ChatGPT ?)

Je teste les GPTs sur COPILOT ! (mieux que sur ChatGPT ?)

🎙 Ludo Salenne 👥 267K 📅 March 12, 2024 ⏱ 22 min 👁 18K 📄 tutorial 🧭 2026-08-21
Available in: English (current) Français

Keywords

Copilot ProGPT BuildercomparisonAI toolsproductivity

Summary

In this video, Ludo Salenne explores the newly available GPTs feature on Microsoft Copilot Pro. He starts by accessing the GPTs section and testing a default GPT (Vacation Planner) to see how it works. Then, he creates a custom GPT from scratch using the GPT Builder, interacting with it to define the name, instructions, and behavior. He demonstrates the configuration interface, noting the absence of conversation starters and actions compared to ChatGPT. He also tests the ability to upload files and enable web browsing. To compare performance, he recreates his existing ‘Copy Boss’ GPT on Copilot by copying the instructions from ChatGPT, then tests both versions with the same email optimization task. He observes that Copilot’s GPT responds much faster but provides slightly less relevant suggestions, while ChatGPT is slower but more accurate. He highlights the ability to export results to Excel or Word as an advantage of Copilot. The video concludes with the creator expressing surprise at the stability and speed of Copilot’s GPTs, and encourages viewers to test and share their experiences.

174 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video provides a practical, hands-on demonstration of using GPTs on Copilot Pro, which is valuable for users considering the feature. The argumentation is based on personal testing and observation, with a clear comparison of speed and output quality between Copilot and ChatGPT. The creator acknowledges the limitations of Copilot’s GPT builder (missing conversation starters and actions) and notes that while Copilot is faster, ChatGPT may offer more relevant suggestions. The reasoning is straightforward and supported by the demonstrated examples.

Scientific Rigor, Source Quality, Title Accuracy

The video is a tutorial and personal review, not a scientific study. The creator does not cite external sources, but provides links to his own resources and previous tutorials. The title accurately reflects the content. The creator’s methodology is transparent: he tests a default GPT, creates a custom GPT, and compares it with an existing GPT on ChatGPT. However, the evaluation is subjective and based on a single test case, which limits the generalizability of the conclusions.

172 words

Title / Content Match

The title accurately reflects the content: the creator tests GPTs on Copilot, compares them with ChatGPT, and shares his findings.

Quality & Reliability

6/10

The video is a hands-on tutorial and comparison of GPTs on Microsoft Copilot Pro versus ChatGPT. The author demonstrates the process live, but the evaluation is subjective and based on personal experience. No external sources are cited beyond the creator's own resources and previous tutorials.

Chapters

Cited Sources

Concurring Sources

Contribution & Novelties

This video provides an early hands-on look at the GPTs feature on Microsoft Copilot Pro, highlighting its capabilities and limitations compared to ChatGPT. It offers practical insights for users who want to create and use custom GPTs on Copilot, including the speed advantage and the missing features. The creator also shares his experience with the GPT Builder interface and the configuration options.

Pour aller plus loin :

  • Microsoft Copilot — Official page for Microsoft Copilot, where users can access the GPTs feature.
  • OpenAI GPT Builder documentation — Official documentation on creating GPTs, which may be useful for understanding the underlying technology.
  • GPT-4 technical report — Research paper on GPT-4, the model behind both ChatGPT and Copilot, providing technical details.

119 words

Radar Profile

The radar chart shows a balanced profile with moderate scores across all dimensions. The video provides a good amount of information and demonstrates practical usage, but the technical depth is limited and the reliability is based on personal experience rather than external sources.

Reliability 6/10