Devenir 100% EXPERT : ChaGPT 5.2 ! Pourquoi TOUT a changé.

Devenir 100% EXPERT : ChaGPT 5.2 ! Pourquoi TOUT a changé.

How I unlock GPT 5.2 to 100% with this method! It's INCREDIBLE!

by Parlons IA

Applied Sciences & Engineering AI — Models & Research UYQArtificial intelligence (AI)UYQFGenerative artificial intelligence
🎙 Parlons IA 👥 17K 📅 December 14, 2025 ⏱ 31 min 👁 4K 📄 tutorial 🧭 2026-09-08
Available in: English (current) Français

Keywords

ChatGPT 5.2reasoning effortverbosityagentic AIfinancial analysis

Summary

This tutorial by ‘Parlons IA’ aims to teach viewers how to use ChatGPT 5.2 effectively for professional tasks, moving beyond simple prompts. The creator criticizes typical influencer tests as inadequate, arguing that they don’t assess real-world utility. He introduces key features like ‘reasoning effort’ (from none to extra high) and ‘verbosity’ controls, which allow users to adjust the model’s depth and length of responses. He demonstrates practical applications, including creating a presentation over 40 minutes and performing a detailed financial analysis of a fictional company, highlighting the model’s agentic capabilities. The video also discusses the model’s training on professional roles (e.g., lawyers, accountants) and compares it to Gemini 3 on ARC-AGI 2 benchmarks. The creator promotes his own training courses and emphasizes the need for structured prompts and workflows to unlock the model’s full potential, contrasting this with the ‘poor’ prompts used by other influencers. The tutorial includes a promotional segment for his formation and ends with a call to action for subscriptions.

163 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video provides a practical, hands-on approach to using ChatGPT 5.2, demonstrating advanced features like reasoning effort and verbosity controls. The argumentation is persuasive, contrasting ‘poor’ prompts with a more structured, workflow-oriented approach. However, the claims about the model’s capabilities and training are largely anecdotal and lack external verification. The creator’s argument that most users are ’not good enough’ to use the model effectively is a strong rhetorical point, but it also serves to promote his paid courses. The demonstrations, particularly the financial analysis, are impressive and suggest a real utility for the model, but the lack of reproducible details (e.g., exact prompts, data) limits the scientific value.

Scientific Rigor, Source Quality, Title Accuracy

The video does not cite any external sources or studies. The creator mentions OpenAI’s documentation and ARC-AGI benchmarks but provides no links or specific references. The description contains links to his own website, blog, and social media, but these are promotional rather than scientific. The title accurately reflects the content, which is a tutorial on using ChatGPT 5.2. The video’s rigor is low due to the absence of verifiable sources and the reliance on personal claims. The creator’s expertise is not formally established, and some statements (e.g., about model training) are unsubstantiated. The video includes a promotional segment for his training courses, which is a conflict of interest but does not necessarily invalidate the technical advice.

239 words

Title / Content Match

The title accurately reflects the content: a tutorial on using ChatGPT 5.2 to its full potential, with a focus on professional applications.

Quality & Reliability

5/10

The video is a tutorial with practical demonstrations, but it lacks verifiable sources and relies heavily on personal claims and promotional content. The creator's expertise is not formally established, and some statements (e.g., about model training) are unsubstantiated.

Chapters

Cited Sources

  • Parlons IA - Formations — The creator's website offering AI training courses, mentioned in the video and description.
  • Blog Medium — The creator's blog, linked in the description, possibly containing additional content.
  • Chaîne Dailymotion — Alternative video platform for the creator's content.
  • Podcast Spotify — The creator's podcast, linked in the description.
  • SEO Agent IA — A tool for SEO, mentioned in the description with a discount code.

Concurring Sources

  • OpenAI Documentation — The creator references OpenAI's documentation for features like reasoning effort, but no specific link is provided.

Dissenting Sources

  • ARC-AGI Benchmark — The creator claims Gemini 3 is the only model capable of solving ARC-AGI 2, but this is not verified and may be inaccurate.

Contribution & Novelties

The video’s main contribution is its emphasis on using ChatGPT 5.2 as a professional agent, with a focus on structured workflows and advanced parameters like reasoning effort and verbosity. It provides a practical demonstration of financial analysis and presentation creation, which is more concrete than typical ‘prompt’ showcases. The creator’s critique of influencer-style testing is a useful perspective, highlighting the gap between marketing and real-world utility.

Pour aller plus loin :

115 words

Radar Profile

The radar profile shows a moderate level of information quantity and technical depth, but lower scores for information quality and reliability. This indicates a tutorial that is informative but lacks rigorous sourcing and may contain unverified claims.

Reliability 4/10