Vous avez manqué 50% de puissance Claude ? Voici pourquoi
Haven't Updated Claude? Claude's New Architecture!
Keywords
Summary
168 words
Critical Evaluation
Value of the Information & Strength of the Argument
The video provides some practical tips on enabling and optimizing the ‘Advisor’ feature, which could be useful for users of Claude. However, the argumentation is largely speculative and lacks concrete evidence. The presenter makes bold claims about performance improvements and future models without citing any official sources or benchmarks. The reasoning is often circular, relying on leaks and unverified information. The promotional aspect for his own training course also detracts from the informational value.
Scientific Rigor, Source Quality, Title Accuracy
The scientific rigor is low. The video does not cite any official Anthropic documentation or research papers. The claims about the ‘Advisor’ feature and ‘Claude Mythos’ are based on leaks and speculation, which are not reliable sources. The title is somewhat clickbait, but the content does address the topic. The description includes links to the creator’s own website and social media, but no external sources for the claims made. The video’s credibility is further undermined by the lack of any critical analysis or counterarguments.
173 words
Title / Content Match
The title is somewhat clickbait, suggesting a missed 50% power, but the content does discuss a new 'Advisor' feature that allegedly boosts performance, so it is partially aligned.
Quality & Reliability
5/10
The video presents speculative claims about a leaked model (Claude Mythos) and an 'Advisor' feature, with no verifiable sources or official documentation. The presenter mixes factual-sounding statements with unsubstantiated predictions, and the promotional content for his own training course reduces credibility.
Key Moments
Markers derived by PSI from the transcript: the creator did not define chapters.
- Introduction: Anthropic's new update and the 'Advisor' feature.
- Claim about 35% of new graduates not finding jobs due to AI.
- Explanation of the 'Advisor' system and its benefits.
- Discussion of the leaked 'Claude Mythos' model and its potential impact.
- How to activate the 'Advisor' feature and combine it with 'Opus plan'.
- Introduction of 'Hyperthinking' and 'Ultra Thinking' for complex reasoning.
- Emphasis on the shift to 'architecting' workflows and the need for human supervision.
Cited Sources
- Parlons IA - Formations — Promotional link for the creator's AI training courses.
- Parlons IA - Blog — Blog with additional content from the creator.
- Parlons IA - Podcast — Podcast link for the creator's content.
- Parlons IA - Dailymotion — Alternative video platform for the creator's content.
- SEO Agent IA — Promotional link for an AI tool, with discount code.
Concurring Sources
- Anthropic's official documentation — Official source for Claude features, though not directly cited in the video.
Dissenting Sources
- Apple's study on AI reasoning — The video references an Apple study on AI reasoning, but no specific source is provided, and the claim is used to support the need for the 'Advisor' feature.
Contribution & Novelties
The video’s main contribution is the practical advice on enabling and optimizing the ‘Advisor’ feature in Claude, which may be new to some users. It also introduces the concept of ‘architecting’ workflows, which is a relevant skill for AI adoption. However, the speculative nature of the claims about ‘Claude Mythos’ and the lack of evidence limit its novelty.
Pour aller plus loin :
- Claude (language model) - Wikipedia — Provides background on Claude and its versions.
- Anthropic’s official documentation — Official source for Claude features and updates.
- Agentic AI - Wikipedia — Overview of agentic AI systems, relevant to the ‘Advisor’ concept.
- Prompt engineering - Wikipedia — Foundational knowledge for the prompt injection techniques mentioned.
115 words
Radar Profile
The radar profile shows moderate scores in information quantity and technical level, but low scores in information quality and reliability. This indicates a video that provides some practical tips but lacks scientific rigor and credible sourcing.