Actus IA : ChatGPT, Gemini, Claude... du nouveau PARTOUT !

Actus IA : ChatGPT, Gemini, Claude... du nouveau PARTOUT !

by Ludo Salenne

Humanities, Social Sciences & Thought AI — News & Applications UYComputer scienceUYQArtificial intelligence (AI)
🎙 Ludo Salenne 👥 266K 📅 May 15, 2026 ⏱ 27 min 👁 21K 📄 news review 🧭 2026-08-02
Available in: English (current) Français

Keywords

AI newsGemini IntelligenceGPT-5.5AI memoryAI agents

Summary

In this video, Ludo Salenne returns to his AI news format after a year, covering the latest developments in artificial intelligence. He starts with Google’s announcement of Gemini Intelligence, which integrates AI deeply into Android 17 and other devices, moving from traditional OS to an ‘intelligence system’. He discusses the implications for search, introducing the concept of ‘agentic search’ and advises content creators to optimize for AI agents. He then talks about GPT-5.5, noting its colder tone and reduced hallucinations (52.5% less on sensitive topics), and the enhanced memory features that allow users to see and control what ChatGPT remembers. He highlights the new Codex mobile feature, which lets users command AI from their phones to work on their computers. He also mentions a study showing that warmer AI tends to lie more, and suggests a prompt to counter this. He promotes his training course, then discusses a leaked AI virus called ‘Mythos’ that bypasses two-factor authentication, emphasizing the need for passkeys. He covers Mistral AI’s Work Mode, Claude for Small Business, the rise of AI agents, Claude’s valuation at 900 billion, and the debate on AI killing developer jobs. He touches on the AI Act and ends with a key number and a common mistake to avoid.

208 words

Critical Evaluation

The video provides a broad overview of recent AI news, but its scientific rigor is limited. The creator presents information with personal commentary and promotional elements, which may introduce bias. While he mentions a study on AI warmth and lying, he does not provide the source or details, making it difficult to verify. The discussion on GPT-5.5’s reduced hallucinations is based on press articles, but again lacks specific citations. The segment on the AI virus ‘Mythos’ is alarming but lacks concrete evidence or references. The advice on optimizing for agentic search is practical but not backed by empirical data. The video’s strength lies in its accessibility and practical tips, but for a scientifically rigorous analysis, it falls short. The creator’s promotion of his training course is transparent but may influence the content’s objectivity. Overall, the video is informative for a general audience but not suitable for those seeking in-depth, evidence-based analysis.

151 words

Title / Content Match

The title accurately reflects the content, which covers recent AI news from major companies like Google, OpenAI, and Anthropic.

Quality & Reliability

6/10

The video is a subjective news review with personal opinions and promotional content. It cites some studies and events but lacks detailed sourcing and verification. The creator provides practical tips but the reliability is moderate due to potential bias and lack of primary sources.

Chapters

Cited Sources

Concurring Sources

  • OpenAI GPT-5.5 — Official OpenAI website, relevant to GPT-5.5.
  • Google Gemini — Official Google Gemini page.

External References

Contribution & Novelties

The video synthesizes recent AI news and offers practical tips for adapting to AI-driven changes, such as optimizing for agentic search and managing AI memory. It highlights the trend of AI becoming more integrated into devices and the potential risks of AI-generated malware.

Pour aller plus loin :

76 words

Radar Profile

The radar profile shows moderate scores across all dimensions, with a slight emphasis on quantity of information. This suggests a video that covers many topics but lacks depth and rigorous sourcing.

Reliability 5/10