J'ai combiné ChatGPT, DeepSeek et Grok 3 – LA FOLIE !

J'ai combiné ChatGPT, DeepSeek et Grok 3 – LA FOLIE !

🎙 Ludo Salenne 👥 267K 📅 February 22, 2025 ⏱ 16 min 👁 55K 📄 tutorial 🧭 2026-08-21
Available in: English (current) Français

Keywords

ChatGPTDeepSeekGrok 3ReplitAutomation

Summary

The video presents a method for creating web applications by combining three AI models: ChatGPT (o3) for drafting specifications, DeepSeek R1 for generating a first version of the code, and Grok 3 for optimizing the final version. The author demonstrates the process with a concrete example: a Tetris-style game with footballs. He shows how to use Replit to deploy the application online, and then explains how to automate the entire workflow using the Make platform. The video also includes a promotional segment for the author’s training course on automating ChatGPT. The method is based on the ‘meta-prompt’ approach, which involves using multiple AIs in sequence to improve performance and results.

110 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video provides a practical, step-by-step demonstration of a multi-AI workflow, which is valuable for viewers interested in leveraging AI for rapid prototyping. The argumentation is based on the author’s direct experience, showing the strengths and limitations of each AI model. The method is presented as a way to overcome the weaknesses of individual models by combining their strengths. However, the argumentation is largely anecdotal and lacks rigorous testing or comparison with alternative approaches. The author’s enthusiasm is evident, but the scientific basis for the claimed superiority of the multi-AI approach is not established.

Scientific Rigor, Source Quality, Title Accuracy

The video does not cite any scientific sources or research papers. The sources provided in the description are links to the tools used (ChatGPT, DeepSeek, Grok, Replit) and to the author’s own training and resources. The title accurately reflects the content, which is a tutorial on combining AI tools. The video is a practical demonstration rather than a scientific study, so the lack of citations is not surprising, but it limits the scientific rigor. The promotional nature of the content, including the training course, also affects the objectivity of the presentation.

200 words

Title / Content Match

The title accurately reflects the content: combining ChatGPT, DeepSeek, and Grok 3 to create applications.

Quality & Reliability

6/10

The video is a practical tutorial demonstrating a workflow combining several AI tools. The information is reproducible and based on direct experience, but the scientific rigor is limited by the absence of citations and the promotional nature of the content.

Chapters

Cited Sources

  • ChatGPT — Used for drafting the project brief.
  • DeepSeek — Used for generating the first version of the application.
  • Grok — Used for optimizing the final version of the application.
  • Replit — Used for deploying the application online.
  • Qwen — Mentioned as an alternative to DeepSeek.
  • Formation Automatiser ChatGPT — Promotional link to the author's training course.
  • Ressources IA — Promotional link to free AI resources.
  • Trustpilot Reviews — Link to customer reviews of the author's services.

Concurring Sources

  • Replit — The tool used for deployment, which is a real platform.
  • Make — Mentioned as an automation tool, though not directly linked in the description.

External References

Contribution & Novelties

The video offers a practical, hands-on demonstration of combining multiple AI models in a pipeline to create a functional web application. The main novelty is the explicit workflow: using ChatGPT for specifications, DeepSeek for initial code generation, and Grok for optimization, followed by deployment via Replit. This approach is presented as a way to leverage the strengths of each model. The video also introduces the concept of automating this workflow using Make, which could be valuable for scaling the process.

Pour aller plus loin :

  • Meta-prompting — A technique for structuring prompts to improve AI output, related to the method described.
  • Multi-agent systems — The concept of multiple AI agents collaborating, which is the basis of the workflow.
  • No-code development — The broader context of building applications without traditional programming, relevant to the use of Replit and Make.

138 words

Radar Profile

The radar profile shows a balanced distribution across the four dimensions, with slightly higher scores in information quantity and technical level, reflecting the tutorial's practical nature. The lower scores in information quality and reliability indicate the lack of scientific rigor and reliance on anecdotal evidence.

Reliability 5/10

💬 No comments were provided for analysis.