65. GPT-5.4, Anthropic vs el Pentágono y Perplexity Computer: la semana que no para

65. GPT-5.4, Anthropic vs el Pentágono y Perplexity Computer: la semana que no para

🎙 Horizonte Artificial 👥 252 📅 March 19, 2026 ⏱ 45 min 👁 14 📄 news review 🧭 2026-08-16
Available in: English (current) Français

Keywords

GPT-5.4AnthropicPerplexity ComputerOpenAIAI agents

Summary

The video is a weekly news review by the Horizonte Artificial podcast, hosted by Joaquín and Guikina. They discuss eight major AI news items from the week of March 5, 2026. The main topics include: OpenAI’s release of GPT-5.4 with native computer control and a million-token context, and Sam Altman’s contradictory statements about AI power versus signing a Pentagon deal; Anthropic’s refusal to comply with Pentagon demands, leading to a ban and subsequent attempts to renegotiate; Perplexity’s launch of Perplexity Computer, an agent orchestrator priced at $200/month; Nvidia’s CEO Jensen Huang distancing from OpenAI and Anthropic; NotebookLM’s new cinematic video summaries; and Claude Code’s new voice mode. The hosts also share personal anecdotes about using AI tools, discuss the implications of AI in geopolitics, and compare various AI products. The tone is informal and opinionated, with a focus on the business and political aspects of AI rather than technical details.

150 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video provides a valuable overview of recent AI developments, highlighting key industry moves and geopolitical tensions. The hosts offer their opinions and personal experiences, which adds a practical perspective. However, the argumentation is largely anecdotal and lacks deep analysis or evidence. They often speculate about motivations and future implications without substantiating claims. The discussion is engaging but not rigorous, making it more suitable for casual viewers than for those seeking in-depth analysis.

Scientific Rigor, Source Quality, Title Accuracy

The video does not cite specific sources, and the hosts rely on their own knowledge and speculation. They mention that they have a Telegram channel for additional resources, but no direct references are provided. The title accurately reflects the content, covering the main news items. The hosts acknowledge uncertainty about some details, such as the availability of features, which indicates a lack of rigorous fact-checking. Overall, the scientific rigor is low, and the content is more opinion-driven than evidence-based.

167 words

Title / Content Match

The title accurately reflects the content, which covers the major AI news of the week, including GPT-5.4, Anthropic's Pentagon issues, and Perplexity Computer.

Quality & Reliability

6/10

The video provides a subjective commentary on recent AI news, mixing factual claims with personal opinions and anecdotes. No primary sources are cited, and the hosts acknowledge uncertainty about some details. The information is generally plausible but lacks verification.

Key Moments

Cited Sources

Contribution & Novelties

The video offers a timely and opinionated summary of the week’s AI news, providing a Spanish-speaking audience with insights into major industry moves. The hosts’ personal experiences with AI tools add a practical dimension. However, the content is not original research and relies on publicly available information.

Pour aller plus loin :

  • OpenAI — Official site for GPT-5.4 and other models.
  • Anthropic — Official site for Claude and company news.
  • Perplexity — Official site for Perplexity Computer and other products.
  • NVIDIA — Official site for company announcements.
  • NotebookLM — Official site for NotebookLM features.

94 words

Radar Profile

The radar profile shows moderate scores across all dimensions, with a slight peak in quantity of information. This indicates a video that provides a decent amount of news but lacks depth and rigor. The technical level is low, making it accessible to a general audience, but the reliability is moderate due to the lack of citations.

Reliability 5/10