I loved ChatGPT Desktop until OpenAI gutted it to make room for Codex and Work

Summary

OpenAI integrated ChatGPT desktop with Codex, removing user-friendly features like screenshotting and 'Work With'. This change prioritizes agentic tools and Codex, potentially impacting free/lower-tier users. The browser version remains unaffected.

Key Takeaways

1. OpenAI added agentic tools and removed everyday features.
2. ChatGPT Work arrived, but the ChatGPT desktop app got worse.
3. ChatGPT still works best in the browser for most users.

Statement Breakdown

• Claimed Facts: 40% of statements the article presents as facts
• Opinions: 50% of statements classified as editorial or subjective
• Claims: 10% of statements surfaced for additional reader evaluation

Credibility & Bias Reasoning

Credibility assessment: The article presents a personal experience and opinion, supported by screenshots and comparisons. While the author's claims are detailed, the lack of direct quotes from OpenAI and the subjective nature of the critique limit its objective credibility.

Bias assessment: Disappointed User. The author expresses strong dissatisfaction with recent changes to the ChatGPT desktop app, focusing on the removal of features they found useful. The language used is emotionally charged, indicating a personal negative experience rather than a neutral assessment.

Note: This article offers a user's perspective on recent changes to the ChatGPT desktop app. Consider it a personal review, and seek official OpenAI statements for a balanced view.

Credibility flag: User Experience Focused

Claimed Facts (8)

• This states a factual observation about how ChatGPT can be accessed on a Mac.
• This describes a specific functional difference between the Mac app and browser version.
• This states a capability of the browser version of ChatGPT.
• This identifies a specific feature that was present in the previous version of the app.
• This is a direct assertion about the nature of the updated application.
• This describes a specific observation from the app's 'about' screen.
• This describes the continued presence of certain features in the browser interface.
• This lists specific features that remain accessible in the browser.

Opinions (10)

• This is a personal statement about the author's usage habits.
• This explains the author's reasoning for using both the browser and app versions.
• This expresses the author's personal positive experience with a specific feature.
• This expresses disappointment about the removal of a feature.
• This indicates personal reliance on a feature that has been removed.
• This is a subjective assessment of the value of a removed feature.
• This is a strong, subjective conclusion about the current state of the app.
• This uses highly subjective and negative language to describe the new app interface.
• This is an ironic statement that implies the author is indeed bitter.
• This is a direct expression of personal desire for a removed feature.

Claims (10)

• While presented as fact, the absolute nature ('never') and the specific limitation might be subject to change or nuanced interpretation, making it potentially debatable.
• While the app is heavily based on Codex, calling it 'not the ChatGPT app at all' is an oversimplification and potentially misleading, as it still functions as a ChatGPT interface.
• This is a speculative statement about OpenAI's financial motivations, presented as a direct consequence without concrete evidence.
• This is a generalized statement about data centers and their cost, used to support an argument about OpenAI's revenue needs, without direct relevance to the app's functionality.
• This is a subjective interpretation of the app's underlying structure, potentially oversimplifying the integration.
• This is a personal observation that may not be universally true for all users or all types of history.
• This is a predictive statement based on the author's opinion, presented as a likely outcome for other users.
• This is a subjective statement about the 'value' and 'sameness' of features, which is open to interpretation.
• This is a subjective positive assessment of Codex, which is not directly supported by evidence within this article's context.
• This is a rhetorical question implying a negative and unnecessary action by OpenAI, framed as a certainty.

Key Sources

• David Gewirtz — Author
• ZDNET — Media Outlet
• OpenAI — AI Research and Deployment Company