Gemini 3 Deep Think Reviewed & More AI Use Cases

Gemini 3 Deep Think Reviewed & More AI Use Cases

by Igor Pogany

Humanities, Social Sciences & Thought AI — Models & Research UYQArtificial intelligence (AI)UYQFGenerative artificial intelligence
🎙 Igor Pogany 👥 480K 📅 December 12, 2025 ⏱ 21 min 👁 33K 📄 news review 🧭 2026-09-08
Available in: English (current) Français

Keywords

Gemini 3 Deep ThinkChatGPT 5.2Claude Opus 4.5prompting techniquesAI use cases

Summary

In this video, Igor Pogany reviews recent AI developments, focusing on the release of Gemini 3 Deep Think and ChatGPT 5.2. He compares these models with Claude Opus 4.5, highlighting their strengths and weaknesses. He demonstrates a new prompting technique suggested by Andrej Karpathy, which involves asking the AI to consider what a group of experts would say on a topic, leading to more varied and actionable advice. He also tests the integration of Adobe Photoshop and Express within ChatGPT, showing how users can perform quick edits and then continue in the full app. Additionally, he showcases Google’s Mixboard with Nano Banana Pro, which can turn a mood board into a presentation. The video includes a sponsored segment for Wispr Flow, a voice dictation tool. Finally, he covers quick hits: Grok’s AI-generated ads, an OpenAI enterprise report, and an Anthropic survey on AI usage. Overall, the video provides practical insights and comparisons for AI enthusiasts.

155 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video offers valuable information by testing and comparing the latest AI models and tools, providing practical demonstrations and personal insights. The argumentation is solid, as the creator supports his claims with hands-on examples and references to official sources. He clearly distinguishes between objective facts and subjective opinions, which enhances credibility. The prompting technique from Karpathy is well-explained and tested, showing tangible benefits. The comparisons between ChatGPT, Gemini, and Claude are nuanced, acknowledging each model’s strengths. The inclusion of enterprise reports adds a data-driven perspective, though the analysis remains at a high level.

Scientific Rigor, Source Quality, Title Accuracy

The video demonstrates good scientific rigor by linking to primary sources such as Google’s blog, Adobe’s blog, and official reports from OpenAI and Anthropic. The creator also corrects a mistake from a previous video, showing transparency. The title accurately reflects the content, which is a review of Gemini 3 Deep Think and various AI use cases. The video includes a sponsored segment for Wispr Flow, which is clearly disclosed. The creator’s personal testing and demonstrations add practical value, though some claims could benefit from more in-depth verification. Overall, the sources are credible and the content aligns well with the title.

208 words

Title / Content Match

The title accurately reflects the content, which focuses on reviewing Gemini 3 Deep Think and showcasing various AI use cases.

Quality & Reliability

7/10

The video provides a balanced overview of recent AI model releases and practical use cases, with hands-on testing of several tools. The creator clearly distinguishes personal opinions from factual announcements and includes links to primary sources. However, some claims lack deep verification and the analysis is largely anecdotal.

Chapters

Cited Sources

Concurring Sources

Dissenting Sources

External References

Contribution & Novelties

The video provides a practical overview of the latest AI model releases and demonstrates innovative use cases, such as the integration of Adobe Photoshop into ChatGPT and the use of Google’s Mixboard with Nano Banana Pro. The prompting technique from Andrej Karpathy is a valuable addition, offering a simple yet effective way to elicit diverse expert perspectives from AI models. The video also highlights the competitive landscape among major AI providers, giving viewers a clear understanding of each model’s strengths.

Pour aller plus loin :

138 words

Radar Profile

The radar profile shows a balanced performance across all dimensions, with slightly higher scores in information quantity and quality, reflecting the video's comprehensive coverage and practical demonstrations. The technical level is moderate, suitable for a general audience, while reliability is solid due to the use of official sources.

Reliability 7/10

💬 Sur les 0 commentaires analysés, aucune tendance n'a pu être dégagée.