Ils ont rendu CHATGPT moins bon... Et c'est du GÉNIE.

Ils ont rendu CHATGPT moins bon... Et c'est du GÉNIE.

by Ludo Salenne

Humanities, Social Sciences & Thought AI — Models & Research UYQArtificial intelligence (AI)UYQFGenerative artificial intelligence
🎙 Ludo Salenne 👥 266K 📅 September 2, 2025 ⏱ 19 min 👁 52K 📄 expert opinion 🧭 2026-08-03
Available in: English (current) Français

Keywords

ChatGPTGPT-5OpenAIAI alignmentuser backlash

Summary

The video analyzes the controversial launch of ChatGPT-5, which many users found slower, less creative, and more robotic. The creator argues that this is not an accident but a deliberate strategy by OpenAI. He compares the situation to the ‘New Coke’ fiasco, where Coca-Cola changed its recipe and faced massive backlash. He suggests that OpenAI intentionally ‘sabotaged’ ChatGPT to implement a stricter AI alignment (deliberative alignment) for safety reasons and to cut costs. He also discusses the psychological ‘change curve’ and the fear of AI taking control, referencing a simulated drone incident. The video concludes that OpenAI’s actions are part of a larger plan to control AI and users, and he offers tips to mitigate the perceived decline.

118 words

Critical Evaluation

The video presents a compelling narrative but relies heavily on speculation and anecdotal evidence. The creator interprets user complaints as evidence of intentional sabotage, but this is an unproven claim. He draws an analogy to the New Coke fiasco, which is apt in terms of user attachment, but he fails to consider that OpenAI’s changes might be driven by technical limitations or genuine safety concerns. The discussion of AI alignment is oversimplified and lacks nuance, and the drone simulation anecdote is presented without verification. The video does not cite any official sources or studies, and the links in the description are mostly promotional. While the video is engaging and raises important questions about AI governance, its scientific rigor is limited. The title/content alignment is good, but the speculative nature of the analysis reduces its reliability.

135 words

Title / Content Match

The title is catchy and matches the content, which argues that OpenAI intentionally degraded ChatGPT for strategic reasons.

Quality & Reliability

6/10

The video presents a speculative analysis of OpenAI's strategy behind GPT-5, mixing user anecdotes, historical analogies, and unverified claims. It lacks concrete evidence and relies on rhetorical persuasion, but it does reference some real concepts like the Kubler-Ross change curve and AI alignment.

Chapters

Cited Sources

  • Trustpilot Reviews — Mentioned as a source of customer reviews for the creator's training.
  • Formation ChatGPT — Promotional link for the creator's ChatGPT training.
  • Ressources IA gratuites — Promotional link for free AI resources.

Concurring Sources

  • OpenAI's official blog on GPT-5 — Official OpenAI announcement of GPT-5, which may provide context on the model's capabilities and limitations.

Dissenting Sources

  • OpenAI's official response to GPT-5 criticism

External References

Contribution & Novelties

The video offers a speculative interpretation of GPT-5’s launch, framing it as a deliberate strategy rather than a technical failure. It connects user dissatisfaction to psychological models like the change curve and historical marketing failures. However, it lacks concrete evidence and relies on conspiracy-like reasoning.

Pour aller plus loin :

  • AI alignment — Overview of the field of AI alignment, relevant to the video’s discussion.
  • Kübler-Ross model — The change curve referenced in the video.
  • New Coke — Historical case study of the Coca-Cola reformulation backlash.

86 words

Radar Profile

The radar profile shows moderate scores across all dimensions, indicating a video that is informative but not highly rigorous. The balance between information quantity and quality is average, with a slight emphasis on narrative over evidence.

Reliability 4/10

💬 The comments are predominantly positive, with viewers praising the video's quality and depth. Many express appreciation for the analysis and the creator's style, though a few note the speculative nature of the content.