Deepmind

in DeepMind they use Claude, the rest of Google engineers want to and cannot

by

Things are hectic at Google. In recent months, some DeepMind engineers have had access to Claude Code and Anthropic models, but in many other parts of the company this tool, which is currently considered the best on the market, has been banned. This has caused strong internal tensions in the company, and is also a sign of something worrying: Google’s AI cannot compete with Anthropic’s at the moment. what has happened. Steve Yegge is a software industry veteran who worked for years at Google. Last week posted a viral tweet in which he explained that after speaking with a current manager, he was concerned about the adoption of AI tools in this company. “The bottom line is that Google engineers have about the same adoption of AI as John Deere, the tractor company,” he said. Or what is the same: one of the most cutting-edge technology companies in the world was being anything but cutting-edge in its use of AI. Demis Hassabis gets angry (and a lot). Recent Nobel laureate Demis Hassabis, head of DeepMind, criticized Yegge’s tweet and told him to stop spreading nonsense. “This publication is totally false and is simply clickbait,” he said. Yegge returns to the fray. The engineer spoke in the following days with more current Google employees and discovered more things that he told in another long tweet. According to their conversations, the internal adoption of AI at Google follows a conventional pattern: 20% reject it, 60% use it in a basic way, and 20% take advantage of it intensively. Either you let me use Claude Code or I’m out of here.. But there was something else: some engineers are prohibited from using Claude Code because it is a competing product, but the vast majority do not have access to those tools because the company wants them to use Google’s tools — that is, the Gemini models in Gemini CLI, its alternative to Claude Code. According to Yeggewhen they tried to force DeepMind engineers to stop using Claude Code and Anthropic models, they refused to stop using the tool and threatened to leave the company. Internal sources of the company confirmed the data published by Yegge in a published news on Business Insider. Sergey Brin gets going. This week, The Information public that Sergey Brin has sent an internal memo to DeepMind engineers and researchers and confirms what was already being said. “To win the final sprint, we must urgently close the gap in agentic execution and turn our models into world-class developers,” the Google co-founder wrote. According to this data, an “assault team” has been created with the direct participation of Brin himself and DeepMind’s CTO, Koray Kavukcuoglu. That statement makes it clear that Gemini is below Claude in capabilities, because otherwise there would be no talk of “closing the gap.” Google and fragmentation. It’s not just that Claude Opus and Claude Code are now better than Gemini and Gemini CLI. Es que además esto ha dejado claro que en empresas tan grandes como Google la fragmentación operativa puede llegar a ser un problema grave. That some can use something that others cannot, and that worse tools are forced just because they are their own, can end up generating internal tensions, as has happened. That is what Google now seems to want to remedy so that all its employees row together. Another victory for Anthropic. All this controversy does nothing but favor Anthopic, which has managed to take the lead in this race – rather marathon – of AI. That the engineers at the prestigious DeepMind prefer their AI platform to Google’s own is an unequivocal sign that today Anthropic is ahead for AI experts. Image | Alex Dudar In Xataka | The tech industry is spending billions of dollars on GPUs for AI. 95% are inactive

Google’s new AI generates interactive worlds from a prompt. Deepmind believes it is a step to get the AGI

by

The Google Deepmind team has announced its new AI model to generate interactive worlds. At the end of last year We were surprised with what Genie could do 2 And the new version is an important leap, one that for Google is an advance in the creation of the General Artificial Intelligence or AGIthat which can match the abilities of the best human. Genie 3. It is the new World Model o Deepmind world model. Allows to create interactive worlds for which we can explore, all from a Prompt of text. The previous model was very limited and could only be used for a few seconds, but with Genie 3 Deepmind promises that it can be explored for “several minutes.” In addition, the resolution has improved at 720p to 24FPS. The model is based on Genie 2 and I see 3. It has memory. It is the most important improvement of the new model. The world is generated through ia as we explore it, but if we turn around and look at something we had already seen, it remains the same. We can also change something, such as painting on a wall, and that is kept as we leave it all the time. This did not happen in previous versions and its creators say they did not explicitly schedule it to do that. As explained in an article in TechcrunchGenie 3 is able to remember what he has already generated to train himself, in this way he learns how the world and his physical works. Interactive. It also emphasizes that events can be added with Prompts additional In his article, Deepmind puts several interactive examples such as a meadow in which we can choose if a tractor, a bear, a horse or hot air balloons will appear. They call it “promptable World Events” and also allows you to change aspects such as weather. Why is it important. Worlds models are useful in different scenarios such as the creation of scenarios for real -time games, in education or in the training of AI agents. Google points to it in its blog as a key step to reach the AGI, that upper artificial intelligence that so many companies are trying to get as soon as possible. These worlds can be used as a training field for other AI, also including robots, cases in which simulating real scenarios is a challenge. In the presentation, the Deepmind team explained how they put an agent in a stage that simulated a warehouse and asked him to approach certain elements, such as a green garbage cube. In all the tests he achieved, according to the Deepmind team “the fact that (the agent) is able to achieve this is because Genie 3 remains coherent.” The competition. The largest IA competition, at least at the level of products for the end user, we see it in the chatbots and, to a lesser extent, in video or audio generators. The world models are less popular among the public and there is not a great competition. Nvidia presented cosmos at the beginning of the year And there are some companies like World Labs They offer similar proposals. We would like to finish this text with a link so you can try it, but Genie 3 is only available in Beta for a very limited group of academics. Image | Deepmind In Xataka | Some researchers created a company where all employees were AI agents. They did not make a quarter of the work

Google Deepmind has made possible the dream of some employees: charging without working

by

Large technology companies are immersed in a Armament career for AI. As in any war, The soldiers They are a strategic element for each of the parties. That is why Google Deepmindhe has chosen to pay some employees the full year’s salary without having to go to work. His only purpose: no The competition fiche. Google pays for not working. Although large technological have adopted policies in which the performance of your employeesGoogle has made the decision to pay some of them for doing nothing. Something that remembers the Tactics used by goal During the pandemic, in which some employees denounced that they had no assigned functions and that they were only hired to avoid the competition. According to The published by Business InsiderDeepmind, Google’s artificial intelligence division, has signed non -competence agreements in the United Kingdom that prevent certain strategic employees from working for rival companies or starting own projects for a year. During this time, professionals receive their full salary, but they are not obliged to work or can search for employment in other companies. A lost year. This type of non -competition clauses seeks to prevent the knowledge and acquired skills By employees of a certain company, they benefit their competitors once their employment relationship is over. However, far from being an advantage for workers, this period can become an obstacle to their professional development that does not occur in any other business sector. Although receiving a salary without working might seem ideal, many affected employees consider that these agreements are harmful to their careers. The damage to employees is evident in a constant development sector such as artificial intelligence, where technologies evolve at a frantic pace and, being out of the market for a year, implies losing key opportunities for Learn and grow professionally. Talent here and now. A former employee of Google has declared Business Insider than startups that are developing models They are not willing to wait between six months and one year to hire talent, so employees with non -competition clauses are trapped in labor limbo that, in reality, remains relevance in the sector. Nando de Freitas, Vice President of Microsoft AI, confirmed in a publication In your X account which receives weekly consultations of desperate employees for escaping these non -competence agreements. “Do not sign these contracts. No American corporation should have so much power, especially in Europe. It is an abuse of power that does not justify any end,” wrote the Microsoft manager. Freitas has urged workers not to sign this type of contract and has pointed out senior Deepmind executives as responsible for this situation. Made the law, made the trap. In the United States, the Federal Commerce Commission (FTC) prohibited most agreements of non -competence due to its harmful effects on workers. However, this regulation does not apply in the United Kingdom, where Deepmind has its main headquarters. California labor regulations He relaxed his position in 2024 With respect to the non -competence clauses and accepted those agreements signed by foreign companies. This administrative subterfuge allows Google Deepmind, based in London, implementing these clauses without legal restrictions. So you can include these clauses in the contracts of your Silicon Valley employees. In Xataka | If you are angry with your company, congratulations. Anger makes you more productive In Xataka | The Z generation sharpens the talent retention problem in companies: they do not marry anyone Image | Unspash (Tommao Wang)

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.