Does anyone else find that ChatGPT 5’s “deep thinking” is really bad?

Evaluating ChatGPT 5’s “Deep Thinking”: A User Perspective on Its Strengths and Limitations

As artificial intelligence continues to evolve, users are increasingly exploring the capabilities and limitations of large language models (LLMs) like ChatGPT. Among the recent iterations, ChatGPT 5 has garnered attention for its improved functionalities. However, firsthand experiences reveal nuanced insights into its performance, particularly regarding its “deep thinking” or analytical processing abilities.

Primary Use Cases for LLMs

Many users employ LLMs across various tasks, including:

• Advanced Search: When seeking information that surpasses the complexity of standard Google searches, LLMs serve as sophisticated query engines capable of providing nuanced responses.

• Ideation and Concept Expansion: Rather than mere brainstorming, users find that LLMs assist in elaborating on initial ideas, helping to explore avenues that may not have been immediately apparent.

• Narrative Development: While LLMs are not typically used for full-scale writing or worldbuilding, they excel at aiding users in exploring characters, settings, or story concepts, thus fostering creative development.

• Coding Assistance: Recent improvements have made LLMs significantly more effective in supporting programming tasks, debugging, and code generation.

Assessing ChatGPT 5’s Improvements and Shortcomings

Compared to previous versions, many users acknowledge that ChatGPT 5 shows noticeable advancements in various domains. Responses often exhibit greater coherence, depth, and nuanced understanding. While it may not reach PhD-level sophistication, small yet meaningful details in its replies can be impressive.

Interestingly, some users report persistent quirks, such as its tendency to repeatedly suggest “making a quick sketch,” which may not always align with their practical needs.

Critical Insights into “Deep Thinking” Capabilities

Despite these improvements, a recurring concern pertains to ChatGPT 5’s ability to perform in-depth analysis or thorough research—what some might label as “deep thinking.” Many users have observed that when the model is prompted to analyze complex topics or undertake detailed reasoning, its responses can devolve into verbose streams of incoherent or poorly structured information. These outputs, while not entirely illegible, often lack the logical coherence expected from genuine analytical thought.

As a result, some users have decided to switch to the “ChatGPT 5 Fast” mode, which tends to produce more reliable responses by prioritizing speed over intricate, slow-analysis processes.

Conclusion

While ChatGPT 5 marks a