openai

His name is Jules and seeks to stand up for OpenAi

by

Google qualifies it as an “asynchronous programming agent”, but it is much easier to define Jules as what it is: an AI system that helps you program more and better. But it is also something else. War of “Vibe Coding “. Cursor or Windsurf have become banners of a new fever for the program assisted by AI. That of “Vibe Coding” in which you talk to do the job or ask you to self-refle the code based on pressing the tab repeatedly (“Tab-Tab-Tab”). OpenAi in fact just Buy Windsurf for 3,000 million dollarsand the reason is evident: to conquer the developer community to use their programming tool with AI and not another. And right is where Jules enters. The machines that scheduled for us. Jules was presented in December on Google Labs preliminary with one objective: not to offer only a co -pilot to program or a code self -completed tool, but an autonomous agent who reads the code, understands what he intends to do and starts working to solve the problem. Jules “sneaks up” in your repository. As his rivals, one of Jules’s keys is that he is able to integrate (“sneak”) in your code repository to be able to analyze it and help you with your project. Clona all the code in a virtual machine on Google Cloud, studies that code to understand it and from there you can perform various tasks. For example: Design execution tests Create new features Provide audio -shaped changes Correct errors Update dependencies versions An assistant in the background. Jules does all this while you focus on any other task, and when he ends up “thinking” what he has to do and presents the registration of changes (diff) made. Your code is yours. As explained in Google, Jules is private by default. Not only that: Google’s model will not make use of that code as data for your delivery, and all the data you use are kept isolated in that execution environment that is created when using this powerful tool. Integrated with github. Another of Jules’s key elements is that it works totally transparent with Github, the par excellence code repositories. You will not have to install anything extra, which will make the workflow theoretically perfect. Public beta. Google stressed that Jules is now available for all United States users in the URL Jules.google. That will allow anyone to try it for free (and without limits as long as the use is not exaggerated) during this preliminary phase of development and deployment. They are good news for developers who want to try it, including those who do not live in the United States, who can do so with the traditional solution: use a VPN to “simulate” that they are in the US and then connect to that website. In Xataka | Openai has just launched his new programming agent. The interesting thing is what you can do when nobody looks

Openai has just launched his new programming agent. The interesting thing is what you can do when nobody looks

by

Artificial intelligence (AI) currently presumes a leading place in the world of programming. More and more software developers resort to AI systems to write code, correct errors or automate repetitive tasks. Openai is betting again for this area with Codexhis new agent. It is a prior view phase tool that acts as a virtual collaborator. Its engine is Codex-1, a variant of O3 adjusted to better understand the needs of modern development. Among their promises are to generate more orderly code, follow instructions with greater precision, among other advantages. How does Codex work? Codex is not a simple assistant that suggests fragments of code. It is a software agent that operates in the background from the cloud. Once connected to your account GITHUByou can access your repository, read files, propose changes and execute tasks such as writing new functions, correcting errors or throwing tests. All this does it autonomously and safely, within an isolated environment (a kind of virtual cloud computer) that simulates your development environment. This should not only protect your system, but also allows Codex to execute tasks without affecting your local workflow. While he works, you can continue using your computer normally. The interesting thing is that it does not execute a single action in turn: it can take care of several tasks at the same time. For example, you can ask you to check a part of the code or Look for mistakes In another section of the project. Each task is managed separately and Codex will inform their real -time progress, allowing the user to review them. The tool is designed to adapt to the way of working of developers. In fact, it can be guided by specific files called agents.md, a kind of instruction manual that allows you to indicate what style follow, how to launch the tests or what practices should be respected within the project. Although Codex has just launched in preview, it is not an unfilming experiment. Openai engineers themselves have been using it for months as part of their workflow. So that? For Automize repetitive tasksrename variables, write tests or outline documentation. The system is also being tested in companies such as Cisco, where they seek to accelerate the development of new ideas, and in temporary, which uses it to purify errors, write automated tests and reorganize large code bases. OpenAi’s recommendation after this first round of tests is clear: assign well -defined tasks, launch several in parallel and experiment with different types of requests. Because the key is to find the exact point where the agent can display his full potential. Codex works remotely, but does it within a controlled environment. As we said above, each task that executes takes place in an isolated virtual machine, without direct Internet connection or access to external services. It can only interact with the code that the user provides and with the tools that are pre -installed using a configuration script. In addition, Openai ensures that he has trained Codex to identify and reject instructions oriented to the creation of harmful software, as tools for Hacking or malware. In any case, we must not lose sight of the fact that this proposal remains a product in research phase, with many aspects to improve. And it is still a generative AI: you can make mistakes or misunderstand certain instructions if the context is not well defined. Openai plans to introduce gradual improvements, such as the possibility of interacting with the agent during tasks, receiving more detailed updates or integrating it with tools such as incident managers. If you want to start trying Codex from today, you must keep in mind that it is not yet available for all users. For now, OpenAi has begun to deploy the tool Among the subscribers of the plans PRO (200 dollars per month)Enterprise and Team. According to the company, the users of the plans Plus and Edu will have “soon” access. Images | OpenAI In Xataka | Saudi Arabia has signed a check of 7,000 million dollars for Nvidia. Jensen Huang is now 12,000 million richer

This model that Openai presents as an improvement against GPT-4O would not reach Chatgpt. But now it’s here

by

When OpenAI presented GPT-4.1 Last April he said that the model would be available only through the API. In other words, that your access would be limited to professional users who develop applications. But this has just changed. GPT-4.1 land in chatgpt. The artificial intelligence startup (AI) Hannounced that “By popular request”, GPT-4.1 will be available in its chatbot application. This alternative has begun its deployment among the users of the Chatgpt Plus, Pro and Team payment plans. Those who use the Entertainment and Edu alternatives will have access “in the coming weeks.” And what about Free users. As we have just seen, GPT-4.1, at least for the moment, it will be limited to those who take out the wallet. But Openai is deploying among all, including the free ones, to GPT-4.1 Mini, a more leading version (and less capable). There are more changes. The positions in the model selector are changing. The firm led by Sam Altman has decided to eliminate GPT-4o Mini. The natural replacement is GPT-4.1 mini that we have just spoken. More capable than GPT-4o. At least according to what OpenAI says, GPT-4.1 is more capable than GPT-4O. If we take into account this, this means that the company is opening a new flaggpt model. Stands out in: Ability to follow instructions. In the Benchmark Multichallenge in Scale, a test that evaluates how well a model follows instructions, GPT-4.1 reaches 38.3%, with an improvement of 10.5 points compared to GPT-4O. Long context. In Video-Mme, a test focused on the understanding of long and multimodal contexts, GPT-4.1 marks a new maximum: 72.0% in the “Long without subtitles” category, surpassing GPT-4O in 6.7 points. PRogramation. GPT-4.1 achieves a 54.6% score in Swe-Bench Verified, improving at 21.4 points to GPT-4O and in 26.6 points to GPT-4.5. An alternative to O3. In the scope of programming, Openai says that GPT.4.1 can become an alternative to O3, its most advanced reasoning model available at this time. This is interesting because GPT.4.1 is not a reasoning model per se, but OpenAi seeks to combine the capacities of its standard models with the reasoning models in the future. Images | OpenAI + Photoshop | Levart_photographer In Xataka | OpenAi plans a future IPO. It is the definitive step to become a profit company

OpenAi plans a future IPO. It is the definitive step to become a profit company

by

Openai and Microsoft are renegotiating The terms of your non-dilio. The artificial intelligence startup recently announced an important change in its restructuring plans, and one of the objectives is that of a potential outlet. To the rhythm they are spending moneybetter that they complete that transition. Profit of profit. Last week OpenAi left his original restructuring plans. Its complex structure makes the “commercial” organization, which we know as OpenAI, is controlled by the non-profit organization (Non-Profit), OpenAi, Inc. Altman’s intention was to completely separate himself from the latter and become a company with profit (“for-profit”) traditional. Money and social good. However, They will become a public benefit corporation controlled by the non -profit organization. This type of entity not only seeks to obtain benefit, but also seeks social good. It is the same model that rivals such as Anthropic or XAI have adopted, and will allow OpenAi to offer their investors a business participation in exchange for their investments. And open the doors to an IPO.. Another of the key elements of the restructuring is that it will allow OpenAI as a public benefit corporation to opt for a public offer of actions to go over. That opens the definitive possibility to obtain funds: sell company participations such as any other company quoted in indexes such as Nasdaq. Microsoft Renegocia. Those renewed Openai intentions have caused them to now maintain a renegotiation of the terms of the alliance with Microsoft. The company of Satya Nadella has invested about 13,000 million dollars in Openai – part of them, in the form of resources to train their models – and OpenAi’s plans offer an opportunity to reach new agreements. More access to chatgpt. According to Financial TimesMicrosoft is willing to give part of its participation in that new corporation with OpenAi for profit in exchange for accessing its models and technology beyond 2030. The current agreement ends at that time and covers the access that Microsoft has to Openai’s intellectual property, in addition to a commission for the income that the company obtains for its commercial products, such as Chatgpt Plus. OpenAi is already giant. In FT they reveal that Openai is valued right now at 260,000 million dollars. However, the recent Softbank investment of 40,000 million dollars in Openai has caused that according to their own responsible Its valuation is 300,000 million dollarsas well as Coca-Cola. The IPO can further shoot its assessment, which is already colossal, thanks to its projection and popularity. But this moment is delicate. Analysts agree that Openai needs to complete that transformation to an entity with profit. If not, I would see future financing rounds committed, especially since companies that bet on OpenAi do it logically with the hope of recovering their investment with juicy benefits. They keep burning money. Meanwhile, in Openai they continue to burn money as if there were no tomorrow to train their generative models. The recent investment round led by SoftBank allows you to continue having a maneuvering room, but in the company they make it clear that they do not expect to have benefits until 2029. New member of the Big Techa group sight. By then, yes, its agricultural models and new products promise to make income of up to $ 125 billion, According to The Information. The step to a “For-Profit” structure is intended to endure until then and continue to have access to new funds in the future. It is a risky commitment, but one that can do well can make it a new full rule Big Tech. Image | Microsoft In Xataka | Silicon Valley has an obsession with “Todismo”: they begin by dominating a sector and then wanting to dominate them all

Altman’s plan to attract private investment to OpenAI has stayed halfway. It’s a victory for Elon Musk

by

Openai has conquered a place at the forefront of artificial intelligence (AI) with products such as Chatgpt and GPT-4O. But there is a detail that escapes in many conversations: It is still a startup. It is several times smaller than Microsoft or Google and faces an existential challenge that does not have so much to do with its products, but with money. That challenge has its own name: its peculiar organizational structure. The openai part that generates income, directed by Sam Altmanis controlled by a non -profit organization. This unconventional model has aroused doubts among investors. Altman proposed to restructure the company to attract more private capital and accelerate the path to a General Artificial Intelligence (AGI). The problem is that this plan has just received a blow. A difficult governance to sell. The Board of Directors has decided to maintain control from the non -profit entity, after opening conversations with the general prosecutors of California and Delaware. These authorities monitor the legal status of organizations of this type and could have blocked change. The announcement has made it public Bret Taylor, president of the Board, In an official statement from OpenAI. A victory for Musk. Elon Musk had sued Openai for that attempt at reorganization. He assured that the company had diverted from its initial purpose of developing a safe and oriented the good of humanity. The decision to preserve the original supervision partially reinforces its argument: the structure remains, at least for the moment. New model, but with the same control. Despite the other way around, Openai maintains another important change. It is expected that its commercial organization will operate as a public benefit corporation (PBC). The difference is that, instead of separating itself from the non -profit organization, it will continue to have the last word. Now they are negotiating how that supervision will be articulated, but everything indicates that it will be the non -profit organization that designates the members of the Board of the New PBC. This could hinder future financing rounds. Because in this OpenAi, which has not managed to transform completely, interest is not measured only into dividends, as investors would like. Outstanding image | Sam Altman (X) + Photoshop | Ted Conference In Xataka | Silicon Valley has an obsession with “Todismo”: they begin by dominating a sector and then wanting to dominate them all

Openai remains to have your own browser. Google has just put yours on a tray

by

Google could force you to sell your Chrome browser. The antitrust trial that is being held in the US now goes through its critical stage. The judges already declared that “Google is a monopoly” In the Terrros of Searches and Advertisingand now it remains to know what the consequences will be for this technological giant. The demand for selling Chrome has already generated the interest of a company to which this browser would come as a ring to the finger. Openai would be interested. Nick Turley, product manager at Openai, declared In this judicial process. When asked if in Openai they would be interested in something like that, he replied that “yes, we would be, as would be other companies.” OpenAI wants to be the new Google with its search engine. Sam Altman’s company launched in summer 2024 Searchgptconverted something later In Chatgpt Search. The OpenAi’s search engine is, as perplexityan increasingly striking alternative to the traditional Google search engine, and with it it was clear that the searches are one that OpenAi seeks to dominate. And now also with a chatgpt browser. But it is that in Openai they not only want to have their own search engine, but also their own browser. In November 2024 we learned that the company is working in its own browser that would have integrated chatgpt natively. Developing a browser is a complex task, so if Chrome finally puts on sale, acquiring it and modifying it would be an interesting solution for Openai. We have already seen how Openai Chrome could be. When the company launched its search engine, it did it both with an integrated option at chatgpt.com, and through an extension for the Chrome browser. In Xataka we tried it at that time “I use it since then,” and it is clear that this chrome+chatgpt combination is really striking. Two out of three users use Chrome worldwide. Source: Statcounter Globalstats. “An incredible experience”. Turley highlighted in his statement that having Chrome more integrated in Openai would allow a much better product. “We could offer a really incredible experience” if Chatgpt was integrated into Chrome, he said. “We would have the ability to teach users what they could expect from an AI experience as the protagonist” Chrome is too dominant (and that is great for Openai). Chrome browser market share is currently 66.17% according to figures Statcounter Globalstats. Two out of three users worldwide use it, and only safari (thanks to their integration in iPhone, iPad and Mac) manages to do some shade with 17% quota. Chrome’s imperial domain would solve a fundamental problem for Openai: the distribution and dissemination of Chatgpt to reach more users. Chatgpt would reach everywhere. It is true that Chatgpt’s popularity is remarkable, but Turley himself claimed that there is a big problem with distribution. Although they have an agreement With Apple on iPhonethey have not managed to reach large agreements with Android devices manufacturers. Being able to integrate chatgpt in Chrome would solve the problem with a blow and porrazo and allow its chatbot to be “pre -installed” in hundreds of millions of devices. OpenAi asks for free competition …. The OpenAi manager He also spoke How some of its competitors “They control access points to how people discover products, including ours. People discover them through a browser or an application store. Having real choice capacity drives competitiveness. Users should have the ability to choose.” … for now. It is, of course, an expected speech of someone who at the moment does not have a monopoly and who wants to try to compete with those who do. It is evident that Turley referred to Apple and Google, the two companies that dominate the market and the distribution of products. Before Microsoft did it and took advantage of for example to convert its Internet Explorer browser into the absolute dominator of the market. Did it until The EU forced him to offer other options And until Firefox and finally Chrome won that game. A decisive judgment. After the legal process that declared that Google is a monopoly in the field of searches and that ended a few months ago, now the phase that in the US calls “remedies” is being held, and that focuses on the measures that the Judicial Court must take to solve the problem. This part of the process will last about three weeks, and Google’s future can be huge impacted depending on the decisions taken. A remote possibility. Be that as it may, nothing is much less determined, and an important carambola would have to be produced for OpenAi to end up getting Chrome. To begin with, American justice would have to force Google to definitely sell Chrome. Google would probably appeal, which would lengthen the sentence became effective. And by then Openai may have launched his own browser, but what is certain is that Chrome would also be interested in many other companies, and there would be a struggle and a spectacular bid to get this development. Image | As photography | Techcrunch In Xataka | Understanding Apple Intelligence, Apple’s great model: thus remains in front of others and what will chatgpt use

The kindness with chatgpt is coming out of OpenAi. The “please” and “thanks” have an absurd cost every month

by

Probably these days, doing Scroll In social networks, you have crossed with one of those memes that say everything with an image: a man about to be eliminated by robots is saved at the last moment because, according to one of them, he “always said thanks.” A minimum, almost automatic gesture, which redeems it to machines in the dystopic future and that starts a smile in the meantime algorithm. Obviously, we do not expect robots to take control of the planet. But it is no accident that this idea has become viral. Cinema and literature have been feeding that imaginary between fascination and fear. The funny thing is that, beyond fiction, there is something real in all this: being educated with artificial intelligence (AI) is not free. Our kindness, however routine, costs money. And not precisely little. The functioning of AI is not exactly cheap. When we use an artificial intelligence chatbot, we rarely think about the consumption of resources that each of our requests implies. Chatgpt and other similar systems work thanks to large language models, also known as LLM, which operate in data centers equipped with high performance hardware, such as NVIDIA H100 GPU. The execution of these models, which is technically called inference, requires enormous processing capacity and, therefore, a considerable amount of electricity. To all this we must add the necessary investment to build and maintain these infrastructure, as well as their environmental impact. We know, for example, that many of these centers consume large volumes of water to cool the systems, since this type of computer generates heat. Very hot. There is no consensus on how much each request costs, and the figures vary according to the company. Even so, Sam Altman has given a track. The amazing cost of kindness. A few days ago, a user asked on social networks how much money Openai would have lost just because people add “please” and “thank you” to interact with their models. CEO’s response did not go unnoticed: “Dozens of millions of dollars well spent.” He did not speak of a single interaction, of course, but about the accumulation of millions of friendly messages over time. Click to see the original message in x It is curious that, despite the high energy and economic cost that this represents, Altman considers that it is a justified expense. Should we be kind to an AI? As in any other area, there are people who prefer to go directly to the point and others who choose to maintain a more careful and educated tone. This difference probably also has to do with the way each one communicates in their day to day. And in this context, kindness is not just a matter of manners. A study entitled “Should We respect LLMS? A Cross-Linguistic Study of Prompt Politeness in Academic Performance”Analyzed the role of courtesy in the indications and left some interesting conclusions. Among them, that unfortunate messages tend to generate a worse performance. Although, yes, excess courtesy does not guarantee better results. What seems clear to those who use this type of chatbots daily, and more now that they are able to learn and remember more precision, they can adapt to the user’s tone. If we want a more respectful, more coherent and a little more human interaction, we should probably start by taking that step. Images | X screen capture In Xataka | The hallucinations are still the Achilles heel of the AI: the latest OpenAI models invent more of the account

OpenAI goes for Windsurf for 3,000 million dollars, according to Bloomberg

by

As he has advanced Bloomberg and has confirmed later CNBC citing its own sources, OpenAi is negotiating the purchase of a programming assistance startup to compete directly in the developer tool market. It’s about Windsurf. Windsurf is a startup with a programming assistant similar to COPILOT. And OpenAi is valuing a possible acquisition for 3,000 million dollars. Why is it important. The purchase would be a strategic movement to compete in the software development market in the AI ​​era. Openai would go into direct competition with Microsoft (its largest investor) and Anysphere, manufacturer of Cursora similar tool. This operation tells us something about Openai: not only wants a great general use model, but to move to at least one more concrete and niche product. Between the lines. This purchase can generate tensions. If OpenAi Buy Windsurf, it will compete directly with the suite Microsoft development, especially with Github Copilot, which dominates the code assistant market with AI. It would not be The first recent trenching between Openai and Microsoft. The context. He Rumore Rumore It arrives just after the launch of OPENAI O3 and O4-MODE and the closure of a financing round of 40,000 million. The IA developed tool market is exploiting. The trend Vibe Codingwhich seeks to generate code based on short descriptions, is being one of the fashions of the year. Windsurf competes in this space with cursor (valued at 10,000 million), Replit and Microsoft solutions. With only 40 million dollars in annual income (five times less than cursor), Windsurf seems more a commitment to talent and technology than for the business they already generate. And now what. The conversations are still underway and the agreement is not closed yet, according to Bloomberg and CNBC. If concretized, the shadow of American anti -protection regulators would soon appear. Especially for the complex relationship between OpenAI and Microsoft. For developers, this purchase would mean some questions about possible changes in the prices and integration of Windsurf with OpenAi APIS. In Xataka | Openai’s hypothetical social network does not want to connect people. Want your data to train your AI Outstanding image | Windsurf, Openai, Xataka with Mockuuuup Studio

O4-mini is much more than another model of AI. It is the Tesla Model 3 of OpenAI

by

The last launch of OpenAI, the models O3 and O4-minihe says much more about his strategy than he seems to the naked eye. The greatness of this proposal-especially that of O4-mini-is not in Benchmarks nor in technical capacities, which also; but in the tactic that raises and that Remember a lot to which he made Tesla In a giant: First came Model sa luxury product that demonstrated what was possible, but accessible to few. 80,000 dollars. Then came Model 3surprisingly powerful but at a price that greatly expanded the market. 35,000 dollars, less than half than Model S but maintaining 80% of its key benefits. With O3 and O4-mini, OpenAi has executed a very similar movement for the world of AI. O4-mini costs just 1.1 dollars per million Tokens input and $ 4.4 per million departure. And offers a performance that rivals his older brother: 68.1% in Swe-Bench compared to 69.1% of O3which costs almost ten times more. This price-reference equation is no accident, but the aforementioned strategy that we have seen before. A “economic” product that democratizes premium capabilities and expands its potential market. O4-mini exceeds Claude 3.7 Sonnet (62.3% in Swe-Bench) costing much less. It is the best model in Aime 2024 and 2025 (elite mathematical contests) surpassing even major systems. Tesla analogy is reinforced when we see that O4-mini allows a user of Chatgpt plus ($ 20 per month) access capacities that previously required Pro subscriptions (200 per month) or Enterprise (thousands). And the master coup comes from integration. O4-mini is the first “mini” model capable of using all chatgpt tools: web navigation, run python code, analyze images and generate them. With a single affordable model, Openai is absorbing use cases for which several specialized suppliers had to be paid. Concrete examples: A startup that makes slate diagrams analysis for $ 50 per month, now competes against a system that does the same for $ 20 per month, and also solves mathematics, program and seeks information. Another that sells assisted programming tools for $ 30 per month faces an O4-mini that writes better code and also visualizes data. This explains Why Openai released O4-mini when Initially he had announced that would integrate this technology directly into GPT-5. The competitive pressure of Claude 3.7 Sonnet and Grok 3 He forced them to accelerate and have shown that they are willing to cannibalize their own premium products to maintain their dominant position. This reminds what Apple did with the iPod: it was not afraid to integrate it into the iPhone knowing that in a five years It would evaporate what generated 40% of its income. Advance or die. The key difference between O4-mini and alternatives such as Claude is very clear: while others offer larger models that consume more resources, OpenAi has achieved a compact model with sufficient performance for almost any practical task. In the same way as the Model 3 It did not need autonomy for 600 kilometers when 350 were enough for the majority, O4-mini does not need the last 5% precision when it already solves 95% of everyday problems. It is not just a technological strategy, but of scale economy: With each new user, the marginal cost decreases while the competitive advantage increases. And as Tesla, when the competition tries to react by copying its approach, Openai is already announcing its next movement. Of course, as is happening to Tesla, The Chinese threat In the medium term it is unpredictable. In Xataka | Openai’s hypothetical social network does not want to connect people. Want your data to train your AI Outstanding image | OpenAI, Xataka with Mockuuuups Studio

Log In

Forgot password?

Forgot password?

Enter your account data and we will send you a link to reset your password.

Your password reset link appears to be invalid or expired.

Log in

Privacy Policy

Add to Collection

No Collections

Here you'll find all collections you've created before.