J'ai créé 5000 prompts ChatGPT — Voilà ce que j'ai appris

Value of the Information & Strength of the Argument

The video provides practical, actionable advice for improving ChatGPT interactions, with clear examples and demonstrations. The tips are grounded in the author’s extensive personal experience and are consistent with known prompt engineering principles. The argumentation is logical, explaining the reasoning behind each tip, such as the predictive nature of language models and the impact of system prompts. However, the evidence is largely anecdotal, and the video does not provide systematic comparisons or rigorous testing. The mention of the Google DeepMind paper adds credibility, but the author does not delve into the study’s methodology or limitations. Overall, the value is high for practitioners seeking to improve their use of ChatGPT, but the scientific rigor is moderate.

Scientific Rigor, Source Quality, Title Accuracy

The video cites a Google DeepMind paper (arXiv:2309.03409) and mentions OpenAI’s system prompts, but does not provide detailed references or verify claims. The description includes links to the paper and other resources, but the video itself lacks in-depth source analysis. The title accurately reflects the content, and the video is well-structured with clear chapters. The author’s personal experience is a valid source for practical tips, but the lack of external validation and the promotional nature of some content (e.g., paid course) slightly reduce the overall rigor. The video does not engage with potential counterarguments or alternative perspectives, which limits its scientific depth.

The title accurately reflects the content: the author shares lessons learned from creating 5000 prompts, focusing on advanced tips for better ChatGPT responses.

6/10

The video is a practical tutorial based on personal experience and some references to research (Google DeepMind paper) and OpenAI system prompts. It lacks rigorous scientific methodology and relies on anecdotal evidence, but the advice is generally consistent with known prompt engineering practices.

• 00:00Mes 1ers Prompts
• 02:40Au programme
• 03:51Astuce 1
• 09:13Astuce 2
• 15:54Astuce 3
• 20:18Critère Personality v2
• 21:16Astuce 4
• 23:13Astuce 5
• 25:41Astuce Bonus

• Google DeepMind paper on prompting techniques — Referenced in the video as the source for the 'take a deep breath' tip.
• Formation ChatGPT (paid course) — Promoted in the description as a resource for automating ChatGPT.
• Free AI resources (prompts & GPTs) — Mentioned in the video as a source for free prompts and GPTs.
• Article about Apple AI chip — Linked in the description, but not discussed in the video.
• Tuto pour bien utiliser ChatGPT — Referenced as a previous tutorial on basics.
• Tuto pour rédiger les meilleurs prompts — Referenced as a tutorial on prompt writing.
• Tuto créer des GPTs — Referenced as a tutorial on creating GPTs.
• Tuto ChatGPT Memory — Referenced as a tutorial on ChatGPT Memory.
• Tuto Custom Instructions — Referenced as a tutorial on custom instructions.
• Tuto ChatGPT Vision (GPT-4V) — Referenced as a tutorial on ChatGPT Vision.
• ChatGTP est bridé (System Prompt) — Referenced as a video about ChatGPT's system prompt.
• Test Perplexity AI — Referenced as a test of Perplexity AI.
• Tuto Playground OpenAI — Referenced as a tutorial on OpenAI Playground.
• La méthode de l'Arbre de Pensées — Referenced as a video on the Tree of Thoughts method.

• Google DeepMind paper on prompting techniques — Supports the 'take a deep breath' tip.
• OpenAI Playground documentation — Explains parameters like temperature and top-p, which the video references.

• Critique of 'take a deep breath' technique — Some studies suggest that such phrases may not consistently improve performance across all tasks, and the effect might be due to increased verbosity rather than actual reasoning.