J'ai forcé 3 IA à prédire les 104 matchs de la Coupe du Monde.

J'ai forcé 3 IA à prédire les 104 matchs de la Coupe du Monde.

by Ludo Salenne

Applied Sciences & Engineering AI — Models & Research UYQArtificial intelligence (AI)UYQFGenerative artificial intelligence
🎙 Ludo Salenne 👥 266K 📅 May 31, 2026 ⏱ 30 min 👁 4K 📄 tutorial 🧭 2026-08-02
Available in: English (current) Français

Keywords

prédictionsCoupe du MondeIAsimulationmatchs

Summary

In this video, Ludo Salenne presents an experiment where he asks three AI models (ChatGPT, Gemini, and Claude) to predict the outcomes of all 104 matches of the 2026 FIFA World Cup. He built a custom application using Claude Code to interface with the AI APIs, providing them with a detailed context including team rosters, playing styles, and environmental factors, while deliberately excluding betting odds to avoid bias. The video shows the step-by-step process of setting up the simulation and then reveals each AI’s predictions. Gemini predicts France will win against Brazil in the final, with Haaland as top scorer and Mbappé as best player. Claude also predicts France as champion, but against Spain in the final, with similar semi-finalists. ChatGPT’s predictions are not shown in the transcript but are mentioned in the chapters. The creator emphasizes the educational and entertainment purpose, clarifying that the app is not a betting tool and AI predictions are not reliable. He also discusses the importance of context and avoiding biases in AI interactions. The video includes promotional segments for his AI training and community.

181 words

Critical Evaluation

The video presents an interesting and entertaining application of AI to a popular topic, but it lacks scientific rigor. The creator clearly explains his methodology: he built a custom app to query three AI models via APIs, provided them with a detailed context (team rosters, playing styles, environmental conditions), and deliberately excluded betting odds to avoid bias. This is a commendable approach to reduce some biases. However, the predictions themselves are purely speculative and not based on any statistical model or historical data analysis. The AI models are essentially generating plausible narratives based on their training data, which includes a vast amount of football information, but they have no true predictive power. The creator acknowledges this, stating that the video is for educational and entertainment purposes and that the app is not a betting tool. He also notes that the AI’s confidence scores (e.g., Gemini’s 65%) are likely overestimated. The video’s strength lies in its transparency about the process and its discussion of AI limitations. It also touches on important concepts like prompt engineering, context setting, and API usage. However, the content is not scientifically validated; there is no comparison with actual match outcomes (since the tournament hasn’t happened yet), and the predictions are presented without any critical analysis of their plausibility. The creator does not provide any external sources or references to support the claims made by the AIs. The video is well-structured with clear chapters, and the creator’s explanations are accessible. The promotional segments for his courses and community are clearly separated and do not detract from the main content. Overall, the video is a decent introduction to using AI for speculative predictions, but it should not be taken as a serious scientific analysis.

286 words

Title / Content Match

The title accurately reflects the content: the creator forces three AI models to predict all 104 matches of the 2026 World Cup.

Quality & Reliability

6/10

The video is a tutorial-style demonstration of using AI to predict World Cup matches. It is transparent about the methodology (custom app, API calls, context setting) and acknowledges limitations (not a betting tool, AI predictions are not reliable). However, the content is primarily entertainment and lacks rigorous scientific validation. The creator's own AI-related courses are promoted, which may introduce bias.

Chapters

Cited Sources

External References

Contribution & Novelties

The video demonstrates a practical application of using multiple AI models to simulate a complex real-world event (World Cup) with a custom-built interface. It highlights the importance of context and bias mitigation in AI interactions. The creator’s approach of using APIs and a custom app to standardize queries is a useful technique for comparative analysis.

Pour aller plus loin :

87 words

Radar Profile

The radar profile shows moderate scores across all dimensions, with quantity of information slightly higher than quality and technical level. This reflects a video that provides a good amount of content but lacks deep technical or scientific depth.

Reliability 5/10