Codex Archives - usatoday24

OpenAI wanted to simplify its AI products. You’ve ended up with a salad of ChatGPT, Atlas and Codex

July 11, 2026 by usatoday24

If you have become accustomed to opening the application ChatGPT and always work from the same place, when you update you may find yourself with a quite different experience. OpenAI is reorganizing its products and, with them, the routines of those who had already turned their solutions into tools for daily use. The move promises to concentrate functions, but it also forces us to understand again which application should be opened, which environment corresponds to each task, and which parts of the previous flow continue to work the same. Three main pieces. The desktop application that we knew is now called ChatGPT Classic, while a new ChatGPT application brings together Chat, Work and Codex in the same environment. At the same time, Atlas enters its countdown: the browser will remain available for a few weeks, but OpenAI has set its withdrawal for August 9, 2026. The company thus concentrates more capabilities under a single brand, although during the transition names, applications and functions that are not obvious at first glance will coexist. They are not three versions of the same ChatGPT. Chat continues to be the space to talk, write, search for information, analyze files or resolve specific queries. Work is designed for longer, multi-stage tasks, such as researching a topic, cross-referencing documents, and preparing reports or presentations. Codex keeps the focus on software development and can work with code, repositories, terminals and other technical tools. All three share an application, but each one responds to a different way of working. The new desktop experience. The changes land on both macOS and Windows. Chat access appears in the left panel next to Work and Codex. When selected, the conversation opens within the general environment, although it can be separated into another window. OpenAI thus retains the most recognizable feature of ChatGPT, but makes it a part of a broader product, with a different organization for those who used the previous application. Atlas, in retreat. OpenAI is moving some of its capabilities to the new app’s built-in browser and to a extension for chrome. The architecture also includes a remote browser from which agents can complete tasks. Anyone who used Atlas as their regular browser will have to prepare for the move: the bookmarks they want to keep can be exported to Chrome, while cookies and passwords can be moved to the new application. Open tabs and browsing history are not transferred automatically. There are alternatives outside. Google is taking Gemini to Chrome itself to understand the open page, summarize content, compare information between multiple tabs, and, in supported accounts, complete multi-step actions. The deployment continues to be limited by region and Spain is not currently among the supported territories. Comet by Perplexitymaintains the browser-agnostic formula and is available on Mac, Windows, Android, iPhone and iPad. Perplexity Pro users can also choose between models from companies such as OpenAI, Anthropic and Google. The idea behind the changes. OpenAI wants ChatGPT to stop being just a chatbot and become the gateway to practically its entire ecosystem. Instead of maintaining separate apps for chatting, scheduling, or browsing, concentrate those functions under one brand. The commitment is reminiscent of the evolution of other large technological platforms, which over time have absorbed previously separate tools to reduce fragmentation and turn a single application into the center of the experience. Background. OpenAI wanted to reduce the complexity of the selector with the arrival of GPT-5but encountered users who did not see the models as interchangeable parts. Some depended on GPT-4o for specific jobs; others preferred his creative abilities and his way of conversing. The company backtracked and offered it again to paying users, before permanently removing it from ChatGPT on February 13, 2026. Perhaps in a few months this transition will be easier than it seems today. But the change brings to mind a well-known idea in software design: We don’t just learn to use a tool, we also build habits around it. When these habits are changed suddenly, the adaptation period becomes as much a part of the experience as the new functions. OpenAI’s challenge will be to make the integration easier for the user than the collection of products it aims to replace. Images | OpenAi In Xataka | The new Grok 4.5 and Muse Spark 1.1 have a groundbreaking bet: to make 90% of their rivals at 20% of their price

The great revolution of GPT-5.3 Codex and Claude Opus 4.6 is not that they are smarter. It’s that they can improve themselves

February 13, 2026 by usatoday24

Last week, OpenAI and Anthropic simultaneously launched their new AI models specialized in programming: GPT-5.3 Codex and Claude Opus 4.6. Beyond the improvements they represent in performance or speed, which are truly amazing, both companies also stated something that completely changes the rules of the game: AI models are actively participating in their own development. Or put another way: AI is improving itself. Why does this change matter?. Generative artificial intelligence tools are reaching a high level of efficiency and precision, becoming in a few years from being co-workers for simple and specific tasks to being able to be involved in a good part of a development. According to the technical documentation of OpenAI, GPT-5.3 Codex “was instrumental in its own creation,” being used to debug its training, manage its deployment, and diagnose evaluation results. On the other hand, it is worth highlighting the words of Dario Amodei, CEO of Anthropic, who in his personal blog affirms that AI writes “much of the code” in his company and that the feedback loop between the current generation and the next “gains momentum month by month.” In detail. What this means in practice is that each new generation of AI helps build the next, more capable one, which in turn will build an even better version. Researchers call it the “intelligence explosion,” and those developing these systems believe the process has already begun. Amodei has declared publicly that we could be “just 1 or 2 years away from a point where the current generation of AI autonomously builds the next.” Most people use free language models that are available to everyone and are moderately capable of certain tasks. But they are also very limited, and are not a good reflection of what a cutting-edge AI model is capable of today. In a brief session with 5.3-Codex I was able to draw this same conclusion, since the AI tools that big technology companies use in their development are nothing like the most commercial ones that we have freely available in terms of capabilities. The code-first approach. Initial specialization in programming makes more sense than we think. And the idea of companies like OpenAI, Anthropic or Google that their systems were exceptional by writing code before anything else is linked to the fact that developing AI requires enormous amounts of code. And if AI can write that code, it can help build its own evolution. “Making AI great at programming was the strategy that unlocked everything else. That’s why they did it first,” Matt Shumer, CEO of OthersideAI, said in a publication that has given us something to talk about these days on social networks. Between the lines. The new models don’t just write code: they make decisions, iterate on their own work, test applications as a human developer would, and refine the result until they are satisfied. “I tell the AI what I want to build. It writes tens of thousands of lines of code. Then it opens the app, clicks the buttons, tests the features. If it doesn’t like something, it goes back and changes it on its own. Only when it decides it meets its own standards does it come back to me,” counted Shumer describing his experience with GPT-5.3 Codex. What changes with self-reference. Until now, each improvement depended on human teams spending months training models, adjusting parameters and correcting errors. Now, some of that work is performed by AI itself, accelerating development cycles. Just like share Shumer and referring to METR dataan organization that measures the ability of these systems to complete complex tasks autonomously, the time that an AI can work without human intervention doubles approximately every seven months, and there are already recent indications that that period could be reduced to four. And now what. If this trend continues, by 2027 we could see systems capable of working autonomously for weeks on entire projects. Amodei has spoken of models “substantially smarter than almost all humans in almost all tasks” by 2026 or 2027. These are not distant predictions, since the technical infrastructure for AI to contribute to its own improvement is already operational. And these capabilities are what are really turning the technology industry on its head. Cover image | OpenAI and Anthropic In Xataka | We have a problem with AI. Those who were most enthusiastic at the beginning are starting to get tired of it.

Openai has a problem with the “Codex” brand. These are all the codex that manages

September 17, 2025 by usatoday24

Openai has just launched GPT-5-Codex. The problem is that I already had three more calls exactly the same. Why is it important. This accumulation of identical names converts the choice of tools into a headache. Each “codex” does something different, but from the outside it seems the same multiplied product. In detail. The “Codex” family has these members: GPT-5-Codexthe newcomer. A model that program for hours without supervision. Change speed according to complexity: fast for simple, slow and meticulous tasks for large projects. Codex Cloudthe veteran. It works as a remote programmer. You send you work and return with code finished after a few minutes of solo work. Codex Clithe local assistant. A terminal utility that helps you from your computer. Competes directly with tools such as Claude Code. Codex (2021). The grandfather of the family. Fed the first versions of Github co -ilotbut it is no longer operational. Between the lines. Openai is trying to fix the linguistic mess. Now writes “GPT-5-codex” with scripts to differentiate it, implicitly admitting that the situation has been lacking. The new model reduces the use of resources into basic tasks by 94%, but multiplies by two the processing time in complex projects. Internally it already supervises more code than human reviewers. The background. Openai seems to have developed these tools without central coordination, something similar to what ended up with the pre- models selectorGPT-5. Each team chose “Codex” independently. And now what. The company prepares access via API for its latest model. Meanwhile, it is time to assume that “Codex” is more a business philosophy than a specific product. The lesson: even the most advanced companies can stumble with something as basic as putting names to their creations. Outstanding image | OpenAI In Xataka | We thought that Chatgpt was used mostly to work. Openai herself has just demonstrated otherwise

Log In

Sign In

Forgot password?

Your password reset link appears to be invalid or expired.

Log in

Privacy Policy

Add to Collection

No Collections