New ChatGPT Model & Memory Features Explained (AI News You Can Use)

New ChatGPT Model & Memory Features Explained (AI News You Can Use)

🎙 Igor Pogany 👥 480K 📅 May 8, 2026 ⏱ 11 min 👁 22K 📄 news review 🧭 2026-09-08
Available in: English (current) Français

Keywords

GPT-5.5 InstantChatGPT memoryhallucination reductionpromptingAI news

Summary

In this episode of ‘AI News You Can Use’, Igor Pogany covers the latest updates from OpenAI, focusing on the new default model GPT-5.5 Instant and improvements to ChatGPT’s memory feature. He explains that OpenAI now recommends a goal-based prompting approach, where users specify the desired outcome rather than step-by-step instructions, and demonstrates this with a live test comparing multi-step and outcome-first prompts. The video also highlights a claimed 50% reduction in hallucinations, particularly in medical, legal, and financial domains, and notes improved search result formatting with better structure and conciseness. Additionally, Igor discusses the release of GPT-Realtime-2, a voice agent model with reasoning capabilities, and the general availability of Claude for Microsoft Office, alongside OpenAI’s ChatGPT for Excel and Google Sheets. The video concludes with a brief mention of other minor stories and a screenshot of the research database used for the show.

144 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video provides valuable, actionable information for everyday ChatGPT users, particularly the shift to goal-based prompting and the improved memory control features. The presenter’s argumentation is clear and practical, using a live demonstration to illustrate the differences in prompting styles. However, the claims about model performance are based on personal testing and anecdotal evidence, not rigorous scientific methodology, which limits the strength of the argumentation.

Scientific Rigor, Source Quality, Title Accuracy

The video references official OpenAI announcements and provides links in the description, which adds credibility. The title accurately reflects the content, focusing on the new model and memory features. The presenter’s testing is not scientifically rigorous, but it is transparent about its limitations. The video does not include a discussion of public comments, so no trends are analyzed.

138 words

Title / Content Match

The title accurately reflects the content: the video explains the new GPT-5.5 Instant model and memory features, along with other AI news.

Quality & Reliability

7/10

The video is a well-structured news review with practical demonstrations and references to official OpenAI announcements. However, the presenter's claims about model performance are based on personal testing and anecdotal evidence, not rigorous scientific methodology.

Chapters

Cited Sources

  • GPT-5.5 Instant — Official OpenAI announcement of the new default model in ChatGPT.
  • GPT-Realtime-2 — Official OpenAI announcement of the new voice agent model with reasoning.
  • AI Advantage Club — Promotional link for the presenter's AI training resources.

Concurring Sources

  • GPT-5.5 Instant — Official announcement aligns with the video's claims about the new model.

Contribution & Novelties

The video offers a practical, user-centric perspective on the latest OpenAI updates, particularly the shift to goal-based prompting and the enhanced memory control features. It provides a clear demonstration of how these changes affect everyday usage, which is valuable for non-technical users.

Pour aller plus loin :

80 words

Radar Profile

The radar profile shows a balanced score across all dimensions, with slightly higher scores in quantity of information and global reliability, indicating a well-rounded and informative video.

Reliability 7/10