Google Deepmind and FC «Liverpool» have created a football coach based on artificial intelligence
Google DeepMind and Premier League club «Liverpool» have created a prototype of the TacticAI artificial intelligence model for football training. The DeepMind lab, which has previously used AI for complex board games such as Go, notes that AI can be trained on patterns that can be seen on sports fields.
Liz Reid has been appointed to head Google's search division — she previously managed search products with artificial intelligence
Google Search has a new head. This business is headed by Liz Reid, who has been working at Google for more than 20 years. Most recently, she was the head of artificial intelligence-based search, known as Search Generative Experience (SGE).
Nvidia AI can turn game characters into chatbots - NPCs with «unique» interaction
During the Game Developers Conference, Nvidia showed how developers can use its artificial intelligence tools «digital human» to voice, animate, and create dialogues for video game characters.
OpenAI boss Sam Altman says the world «doesn't need another copy of Google» because it's «boring»
Instead, developers should focus on creating new ways for people to search and analyze information.
Nvidia launches platform with generative AI to create humanoid robots
Generative AI will help robots perform actions based on input — a combination of text and speech, video, «human demonstrations» and previous experience.
Roblox launches AI-based tools to improve 3D content creation
Roblox has introduced two new AI-based tools aimed at facilitating the creation of 3D content on its platform. At the Game Developers Conference in San Francisco, the company demonstrated Avatar Auto Setup and Texture Generator, which promise to significantly speed up the process of creating avatars and texturing objects.
NVIDIA unveiled the Blackwell architecture, B200 GPUs, NVLINK 7.2T interface, B200 NVL72 servers, and supercomputers based on it
At the GTC March 2024 Keynote event, NVIDIA introduced the Blackwell chip architecture and B200 GPUs based on it. The new GPUs with AI capabilities will replace the retiring H100 and H200 in data centers. Grace Blackwell GB200 chips will combine the Grace architecture with Blackwell.
OpenAI to launch prepaid billing API on March 25
Instead of monthly billing, the company will provide a certain number of credits for using its API or Playground.
Mark Gurman: Apple is in talks with Google to integrate Gemini AI model into iPhone
Apple has also recently discussed the prospects of cooperation with ChatGPT developer OpenAI.
Elon Musk «declassified» the source code of his chatbot Grok — without training data yet
From now on, developers and researchers will be able to use the architecture on which Grok operates in their own projects.
Larry Ellison and Elon Musk team up to apply artificial intelligence to agriculture
Musk’s SpaceX and Ellison’s Oracle will create a new mapping program that will display the types and volumes of agricultural products in the United States, reports Business Insider
US border guards to use X-rays and artificial intelligence to detect drugs
Currently, «high-tech machines» (which are essentially very powerful X-rays) for detecting drugs are being installed at the borders, and their work should be accelerated by artificial intelligence in the future.
GPT-4 artificial intelligence is able to play Doom and does not hesitate when shooting at virtual objects in the form of people
Microsoft scientist and York University researcher Adrian de Winter recently decided to test GPT-4 by running Doom without prior training. The results of this experiment concluded that GPT-4 can play Doom with a certain degree of efficiency without prior training. Although this requires the GPT-4V interface to see the AI and the addition of…
Tesla Optimus can't do that. The Figure AI robot received voice, hearing and vision thanks to OpenAI technologies
It seems that Musk has a one more reason to hate OpenAI.
The incident with the Saudi humanoid robot shows that machines will not be able to take over the world yet
Saudi Arabia’s first «male» humanoid robot has caused a stir after touching a female reporter inappropriately at an artificial intelligence conference last week.
Viral photo of «Hollywood Mountain» in California turned out to be a fake generated by a neural network
Perhaps you’ve come across a stunning bird’s-eye view of «Hollywood Mountain» in California on Facebook or BlueSky? It turns out that it was completely generated by artificial intelligence — in fact, there is no place with that name.
Amazon offers sellers to generate a product page based on its website using artificial intelligence
Sellers on Amazon will soon be able to create product pages automatically using artificial intelligence. The marketplace is releasing a new AI generative feature that takes information from a seller’s external website and then generates a product page on Amazon with a written description and images.
The EU Parliament has approved the AI Act — «the world's strictest» law on artificial intelligence regulation. What's next?
After reaching political consensus in December, EU lawmakers approved by a majority vote (523 — «for») a landmark document that will regulate AI and require transparency from suppliers, reports The Verge.
Google DeepMind has created SIMA, an AI model that plays games. They don't know why — yet.
Google’s DeepMind lab has unveiled SIMA — a learning artificial intelligence model that will play games for you. It’s customized to play more like a human than a superhuman artificial intelligence. SIMA, which stands for Scalable Instructable Multiworld Agent, is currently just a research experiment.
Sign language interpreter, anti-cyberbullying, educational applications: Ukrainian schoolchildren presented their artificial intelligence projects
This year, Ukraine is participating for the first time in the international Teens in AI hackathon, which takes place in more than 40 countries. Last weekend, the national stage of the competition took place, where 10 teams of schoolchildren presented their ideas on how to use AI to solve global problems.
Midjourney is testing a function that will allow «to save» the image of a given character to generate future images
From now on, users can provide the neural network with a reference image of a person and set up the generation so that the next image looks the same.
«Wild disappointment»: tech manager says the internet is stagnating», just like space exploration after the 1970s
Tuesday marked the 35th anniversary of the moment when renowned computer scientist Tim Berners-Lee presented his vision of what would become known as the «World Wide Web». At the same time, Akamai’s Chief Technology Officer Robert Blumoff believes that after a breakthrough and years of development, the Internet is now stagnating.
AI large language models (LLMs) become «more covertly racist» after human intervention
From the very beginning, it was clear that large-scale language models (LLMs) like ChatGPT absorb racist messages from the millions of Internet pages they are trained on. Developers have responded to this by trying to make them less toxic. But new research shows that these efforts, especially as the models get bigger, only serve to stifle racist…
Larry Ellison is $15 billion richer after Oracle shares soar 12% to record high
Today, Allison is one of the five richest people in the world in of the Forbes rating with a net worth of $157.3 billion.
Midjourney bans Stability AI employees after data theft attempt
At the beginning of the month, the service faced a 24-hour outage — allegedly due to «botnet activity» two accounts associated with a competitor company, Stability AI.
How do you like this, OpenAI? Elon Musk to open source his Grok chatbot
Last week, Musk has filed a lawsuit against OpenAI with accusations of abandoning the original mission «to develop AI for the benefit of humanity» in favor of making money.
Breakthrough in matrix multiplication should lead to faster and more efficient AI models
Computer scientists have discovered a new way to multiply large matrices faster by eliminating a previously unknown inefficiency, leading to the largest improvement in matrix multiplication efficiency in a decade. This could eventually speed up artificial intelligence models that rely heavily on matrix multiplication.
Three writers to sue Nvidia for illegal use of their books for AI training
The class action lawsuit, filed Friday night in a San Francisco federal court, says Nvidia used the copyrighted books to train its new NeMo platform, created by the chipmaker as a «fast and affordable way to implement generative AI».
Marilyn Monroe joins the list of dead celebrities who were «resurrected» without consent using artificial intelligence
In recent years, there has been a lot of discussion about the practice of using digital images of celebrities without their consent, both after their death and during their lifetime. For example, Zelda Williams, filmmaker and daughter of Robin Williams — last year criticized the «disturbing» the trend of people trying to recreate the voices and…
Updated: Former Activision Blizzard’s boss Bobby Kotick wants to buy TikTok, — The Wall Street Journal
He has allegedly already presented the idea to potential partners, including OpenAI CEO Sam Altman.
GPT4 is recognized as the best AI model in the chatbot arena
The LMSYS Chatbot Arena service has updated the rating of large language models of artificial intelligence. It allows you to evaluate which models are currently doing the best job.
Microsoft confirms Surface and Windows AI presentation — on March 21
Microsoft has announced an online presentation of Surface, Windows, and Copilot that will take place in two weeks. The event New era of work will take place on March 21. It is aimed at Microsoft business customers and will feature new devices Surface Pro 10 and Surface Laptop 6 for business, as we reported earlier this week, with a focus on the…
Facebook uses artificial intelligence to improve video recommendation algorithm
Facebook CEO Tom Ellison announced that Meta is changing the algorithm for video recommendations in Reels, Groups, and Facebook’s main feed, using artificial intelligence to improve the recommendation algorithm. The world’s largest social network has already migrated its short video recommendation engine Reels to the new one and plans…
Anthropic «forgot» to mention in the Claude 3 announcement that they compared the model to the «old» GPT-4, not the GPT-4 Turbo
The company, founded by several former OpenAI employees, promoted its language model as «better than ChatGPT», but there is a nuance.
Microsoft will unveil Surface Pro 10 and Surface Laptop 6 at the end of the month. The company calls them «AI PCs».
Microsoft plans to unveil new Surface devices in late March. Sources familiar with the plans said that the company will unveil the Surface Pro 10 and Surface Laptop 6, which are aimed at business and commercial customers, on March 21. It calls them AI PCs («AI PC»).
OpenAI says Elon Musk wanted to merge it with Tesla for «absolute control»
The ChatGPT developer has rejected the billionaire’s claims outlined in a lawsuit last week. Musk claimed that OpenAI was allegedly focused on making money and forgot about the primary goal — development «of artificial intelligence for the benefit of humanity».
Google has «definitely messed up» Gemini image generation ─ Sergey Brin, co-founder of the company
Google co-founder Sergey Brin met with entrepreneurs at AGI House, where developers were testing the Google Gemini artificial intelligence model.
ChatGPT can now read responses to you — in 37 languages (including Ukrainian)
ChatGPT voice responses can be activated by clicking the Read Aloud button (the speaker icon in the chat) — the feature is already available in the mobile and web versions for GPT-4 and GPT-3.5.
Anthropic Introduces New Claude 3 AI Model: Beats ChatGPT and Gemini
Anthropic, an artificial intelligence company founded by several former OpenAI employees, says its new Claude 3 family of artificial intelligence models performs as well or better than competitors from Google and OpenAI. Unlike previous versions, Claude 3 is a multimodal model capable of understanding text and image input.
Super Mario by AI: Google DeepMind's new model generates games in the style of classic 2D platformers from scratch
Following Sora by OpenAI, which generates digital virtual worlds in the style of Minecraft Google’s DeepMind generative model has distinguished itself with its own talent for creating games.
Nvidia CEO believes that artificial intelligence will be able to pass human tests in 5 years
Nvidia CEO Jensen Huang said that artificial general intelligence could appear in five years. It will be able to pass human tests.
Experimental AI «worm» Morris II infects users through email clients and self-repairing prompts
A team of researchers has created an AI-based artificial «worm», a malware that can steal data, spread other malware, and spam others via email. A «worm» called Morris II has been developed and successfully operates in a test environment using popular LLMs.
Wikipedia no longer considers CNET a «reliable source» — after scandal with AI-generated articles
Amid the AI boom in 2022, CNET started using the technology to write some articles — and eventually faced criticism and removed part of the materials due to a large number of errors and plagiarism.
Musk sues OpenAI and Sam Altman over «AI threat»
The lawsuit says that OpenAI has allegedly abandoned its original mission of developing artificial intelligence for the benefit of humanity, instead focusing on making a profit.
A scientific journal retracts an article with an AI-generated illustration of a rat «with a disproportionately large genital organ»
An open-access article that contained bizarre, artificially intelligent-generated images of rat genitals, unclear graphics, and captions
Adobe introduces generative artificial intelligence for music creation
Adobe has introduced Project Music GenAI Control, a platform that can generate audio from text descriptions (e.g., «happy dance», «sad jazz») or a reference melody and allows users to customize the results within a single workflow.
Microsoft is working with Nvidia, AMD, and Intel to improve upscaling in PC games. The joint project is called DirectSR
Microsoft has introduced a new Windows API designed to enable game developers to seamlessly integrate AI-powered ultra-high-resolution scaling features from Nvidia, AMD, and Intel.
Apple invests heavily in artificial intelligence and prepares several announcements in this area - Tim Cook
Apple CEO Tim Cook said that his company is investing in artificial intelligence. The company sees «the incredible disruptive potential for generative artificial intelligence, so we are now investing heavily in this area,» Cook said at Apple’s annual shareholders meeting.
Honor is developing a smart ring to compete with Samsung Galaxy Ring
China’s Honor is working on a smart ring, the company’s CEO George Zhao said the day after Samsung officially announced its Galaxy Ring.
Tumblr and WordPress to follow Reddit in selling user data for AI training
Automattic, which owns Tumblr and WordPress, is in talks with Midjourney and OpenAI to sell data from user posts to train artificial intelligence systems.
Spelling error report
The following text will be sent to our editors: