¡El Futuro YA ESTÁ AQUÍ! ChatGPT, Atlas y las Skills de Claude

¡El Futuro YA ESTÁ AQUÍ! ChatGPT, Atlas y las Skills de Claude

🎙 Codemancers - Inteligencia Artificial 👥 2K 📅 October 27, 2025 ⏱ 65 min 👁 236 📄 news review 🧭 2026-08-15
Available in: English (current) Français

Keywords

ChatGPT AtlasClaude SkillsAI browseragent modeprompt injectionAnthropicOpenAI

Summary

In this episode, the hosts discuss two major AI news items: OpenAI’s ChatGPT Atlas browser and Anthropic’s Claude Skills. They provide a live demo of Atlas, highlighting its integration with ChatGPT, its agent mode for automating web tasks, and a security vulnerability (prompt injection) they discovered. They also analyze Claude Skills, arguing that its deterministic code-based approach is more powerful than MCPs. The hosts debate whether Atlas’s agentic approach or Claude’s code-based skills represent the future of AI tools. They conclude that while Atlas shows potential, it is not yet ready for daily use due to slowness, cognitive load, and privacy concerns. They also mention a ‘Razzie’ award for the most absurd code of the week.

116 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video provides valuable hands-on insights into two emerging AI tools. The hosts demonstrate real-world usage of ChatGPT Atlas, including its strengths and weaknesses, and offer a critical perspective on its agent mode. They also present a compelling argument for Claude Skills, emphasizing the advantages of deterministic code over MCPs. However, the argumentation is largely based on personal experience and opinion, lacking empirical data or comparative benchmarks. The hosts do not provide evidence for their claims about Claude Skills’ superiority, and the discussion is somewhat one-sided.

Scientific Rigor, Source Quality, Title Accuracy

The video does not cite specific sources or references. The hosts mention the tools themselves but do not link to official documentation or research. The title accurately reflects the content, focusing on ChatGPT Atlas and Claude Skills. The discussion is informal and lacks scientific rigor, with no mention of peer-reviewed studies or official announcements. The hosts’ claims about security vulnerabilities are based on their own testing, but they do not provide detailed technical evidence.

175 words

Title / Content Match

The title accurately reflects the content, which focuses on ChatGPT Atlas and Claude Skills as key AI developments.

Quality & Reliability

6/10

The video is a podcast-style review of two AI tools (ChatGPT Atlas and Claude Skills). It includes hands-on testing and personal opinions, but lacks formal citations or references to official documentation. The hosts demonstrate practical use and identify security issues, but the analysis is subjective and not peer-reviewed.

Chapters

Cited Sources

Contribution & Novelties

The video offers a practical, hands-on review of ChatGPT Atlas and Claude Skills, highlighting real-world usability and security concerns. It provides a unique perspective on the potential of AI browsers and the advantages of code-based skills over MCPs. The hosts’ live demo of Atlas’s agent mode and their discovery of a prompt injection vulnerability add original value.

Pour aller plus loin :

94 words

Radar Profile

The radar chart shows a balanced profile with moderate scores across all dimensions. The video provides a reasonable amount of information and technical depth, but lacks rigorous sourcing and formal validation. The overall reliability is moderate, reflecting the subjective nature of the review.

Reliability 5/10