Is this the YEAR or DECADE of AI Agents & Agentic AI?

Is this the YEAR or DECADE of AI Agents & Agentic AI?

🎙 Martin Keen 👥 1.8M 📅 November 24, 2025 ⏱ 13 min 👁 29K 📄 expert opinion 🧭 2026-08-06
Available in: English (current) Français

Keywords

AI agentsagentic AIcoding assistantstravel bookingIT support

Summary

The video, presented by Martin Keen from IBM Technology, addresses the debate on whether AI agents are ready for mainstream use now or only in the coming decade. It contrasts the optimistic ‘year of AI agents’ with the more cautious ‘decade of AI agents’ perspective, citing Andrej Karpathy’s view that current agents struggle with basic tasks. The presenter examines three use cases: coding assistants, travel booking, and automated IT support. Coding assistants are highlighted as a success story because they operate in structured environments with clear rules and immediate feedback, leveraging pattern matching rather than human-level reasoning. Travel booking, a common demo scenario, works for simple cases but fails on edge cases, UI navigation, multimodal understanding, and continual learning. Automated IT support is presented as an aspirational use case that requires high reliability in computer use, handling unique user setups, and learning from feedback, which current models lack. The conclusion is that we are in the year of AI agents for narrow, well-defined tasks, but the decade for broader, real-world applications. The video is informative and balanced, though it lacks external citations and relies on the presenter’s expertise.

188 words

Critical Evaluation

The video provides a clear and structured analysis of the current state of AI agents, focusing on three representative use cases. The presenter, Martin Keen, an IBM Technology expert, demonstrates a good understanding of the technical capabilities and limitations of AI agents. The argumentation is logical: he defines four key capabilities (intelligence, computer use, multimodal, continual learning) and systematically evaluates each use case against them. This framework is effective for comparing the maturity of different applications.

The strength of the video lies in its balanced perspective. It acknowledges that coding assistants are already providing significant utility, while travel booking and IT support are not yet reliable enough for full autonomy. This nuanced view avoids both hype and undue skepticism, which is valuable in a field often characterized by extremes.

However, the video has some weaknesses. It relies heavily on anecdotal evidence and personal experience rather than citing specific studies or data. For instance, when discussing travel booking failures, he mentions his own struggles but does not provide concrete examples or statistics. This reduces the scientific rigor of the content. Additionally, the video does not reference any external sources, such as research papers or industry reports, which would strengthen the credibility of the claims. The lack of citations is a notable omission for a technical audience.

The adéquation between title and content is good: the title poses a question that the video directly addresses, and the conclusion is clear. The video is well-paced and accessible, but it does not delve deeply into technical details, which might leave advanced viewers wanting more.

Overall, the video is a valuable overview for those interested in the practical applications of AI agents. It offers a realistic assessment of their current capabilities and future potential. However, for a more rigorous analysis, viewers would need to consult additional sources. The video’s strength is its clarity and structure, but its reliance on opinion rather than data limits its scientific value.

322 words

Title / Content Match

The title accurately reflects the content, which explores whether AI agents are ready for widespread adoption now or in the next decade.

Quality & Reliability

7/10

The video presents a balanced expert opinion from an IBM Technology representative, discussing the current capabilities and limitations of AI agents. It is well-structured and uses concrete examples, but lacks citations to external sources and relies on anecdotal evidence.

Key Moments

Cited Sources

Concurring Sources

Contribution & Novelties

The video provides a clear framework for evaluating AI agent readiness across different use cases, emphasizing the importance of structured environments and feedback loops. It offers a balanced perspective on the hype versus reality, which is valuable for practitioners.

Pour aller plus loin :

77 words

Radar Profile

The radar profile shows moderate scores across all dimensions, indicating a balanced but not deeply technical presentation. The video is informative but lacks rigorous citations, which is reflected in the lower reliability score.

Reliability 7/10