Milestone Technocrat LLC

The explicit Taylor Swift AI fakes, according to Satya Nadella, are "alarming and terrible." Microsoft CEO Satya Nadella has reacted to the issue around artificial intelligence (AI)-generated, sexually graphic spoof photos of Taylor Swift. Nadella describes the spread of nonconsensual simulated nudes as "alarming and terrible," telling interviewer Lester Holt that "I think it behoves us to move fast on this" in an NBC Nightly News interview that will run on Tuesday. Holt asks Nadella how he feels about the internet "exploding with fake, and I emphasise fake, sexually explicit images of Taylor Swift" in an NBC transcript released before the show on January 30. While saying relatively little about them, Nadella's reaction manages to open multiple cans of tech policy worms, which is to be expected given the lack of a clear solution to the problem. There could be a relationship between Microsoft and the phoney Swift images. They originated from a nonconsensual porn-making community on Telegram, according to a 404 Media investigation, which suggests utilising the Microsoft Designer picture generator. Theoretically, the designer won't create photographs of well-known individuals, but AI generators are simple to trick, and 404 discovered that you could circumvent its guidelines by making slight adjustments to the prompts. Even if it doesn't confirm Designer was used to create the Swift images, Microsoft can address this kind of technical issue. However, the process of making false nudes of actual individuals has become incredibly simple thanks to AI algorithms, which is upsetting women who are considerably less powerful and well-known than Swift. Furthermore, it takes more than just forcing large corporations to strengthen their barriers to restrict their output. Despite efforts to make it more difficult, people can teach open tools like Stable Diffusion to produce NSFW images, even if big "Big Tech" platforms like Microsoft are locked down. These generators may be accessed by many fewer people, but the Swift incident shows how far the efforts of a tiny community may go. Other temporary solutions exist, such as social networks restricting the dissemination of nonconsensual material or, reportedly, Swiftie-enforced vigilante punishment against those who disseminate it. Is it considered a "convergence on certain norms"? Currently, though, organising Microsoft's AI division is Nadella's only well-defined strategy. #tech #news #ai #generative #taylor #swift #law #microsoft #organization #community #interview #nadella #convergence #solution #nbc #report #temporary #social #network

3D and VR meetings are now supported by Microsoft Teams. Today, Microsoft Teams will introduce its immersive 3D meetings. With the integration of Microsoft Mesh, the company's mixed reality platform, into Teams, users will be able to convene in virtual environments with or without a virtual reality headset. Mesh feels a lot like a corporate version of AltspaceVR, the social VR platform that Microsoft acquired in 2017 and shut down a year ago. I had the opportunity to test it out inside Microsoft Teams at the end of last year. With a VR headset, 3D Mesh meetings in Teams function best; at the moment, Microsoft only supports Meta's Quest devices. You can take part in 3D virtual meetings, and the amazing spatial audio function allows you to move away from other employees in a virtual environment to simulate having private chats much like you would in an office. With games like bean bag throwing and icebreaker questions for distant coworkers you might not have met before, the 3D surroundings are also customisable. Every basic Mesh feature, including Teams' immersive areas, will be accessible to Microsoft Teams subscribers. It need a Teams Premium licence to set up a personalised immersive environment. The integration of Microsoft Mesh with Teams seems like a logical next step, even though it still serves as a platform for developers to access VR and AR experiences. Concerns about corporate acceptance of these kinds of virtual meetings still exist. Though it's a small subset of the businesses who regularly use Teams, Microsoft claims that Accenture, BP, Takeda, and Mercy Ships are among the enterprises that are now utilising Mesh. Although the Teams integration will be beneficial, Mesh already seems to be moving towards becoming a feature of Teams rather than a stand-alone application. #tech #news #microsoft #vr #meeting #accept #developer #platfrom #experience #headset #business #feature #teams #company

Russian SolarWinds hackers obtained emails belonging to Microsoft's "senior leadership." A few days after declaring a significant update to its software security, Microsoft was the target of a nation-state attack. Today, Microsoft is revealing that it has found evidence of a nation-state attack on its corporate systems, carried out by the same group of hackers supported by the Russian state that carried out the highly skilled SolarWinds operation. Microsoft claims that late last year, some of its senior leadership team members' email accounts were compromised by hackers known as Nobelium. According to a blog post published late on Friday by the Microsoft Security Response Centre, "beginning in late November 2023, the threat actor used a password spray attack to compromise a legacy non-production test tenant account and gain a foothold. Then, using the account's permissions to access a very small percentage of Microsoft corporate email accounts, including members of our senior leadership team and employees in our cybersecurity, legal, and other functions, and exfiltrated some emails and attached documents." It is unclear what other emails and documents were taken in the process, but Microsoft claims that the gang was "initially targeting email accounts" for personal information. Microsoft didn't reveal how long the attackers were able to access its systems; the incident was first discovered last week on January 12. "There was no vulnerability in Microsoft services or products that led to the assault. According to Microsoft, there is currently no proof that the threat actor got access to source code, AI systems, production systems, or client environments. #tech #news #microsoft #security #hacker #email #ai #source #code #client #environment #service #cybersecurity #breach #discovery

An AI Executive Became an AI Defender to Protect Artists According to Ed Newton-Rex, an executive at Stability AI and the former chief AI designer at TikTok, generative AI faces an ethical dilemma. Because of the company's position on training data gathering, he resigned in November. According to Newton-Rex, who has been talking about moral issues in AI research, a lot of people want generative AI models to treat creators fairly. He thinks it would help these creators make better decisions if they had better tools. Fairly Trained is a nonprofit organisation founded by Newton-Rex that offers certification to AI firms who licence their training data. Companies must demonstrate that their training data was either publicly available, specifically licenced for training purposes, provided under an acceptable open licence, or previously owned by the company in order to be eligible for the L Certification programme. The AI sector has its own kind of "fair trade" certifications with this one. Nine businesses, including the music creation platform LifeScore Music and the picture generator Bria AI, have obtained the Fairly Certified certification for artificial intelligence. The top generative AI firm in the world, OpenAI, contends that unlicensed data is a need for developing AI services. But there's a disagreement between Newton-Rex and the first businesses to be approved by Fairly Certified. The creator of Lifescore Music, Tom Gruber, says that it is simple for them to be compliant because the music industry values provenance and rights. Bria CEO, Yair Adato, draws comparisons between Napster and the Pirate Bay and AI models created on unlicensed data. The non-profit levies a $500–$6,000 application fee, dependent on the applicant's business size. Prominent trade associations such as the Association of American Publishers and the Association of Independent Music Publishers have endorsed Fairly Trained, a nonprofit organisation dedicated to standardising AI labelling with ingredient information. But there is still a long way to go until the AI industry abandons its common practice of harvesting training data. The one-man show Newton-Rex is comparable to Adobe's Content Authority Initiative, which tracks the validity of images. In the future, he intends to roll out more certificates, perhaps addressing concerns like pay. Turkewitz thinks the groundbreaking effort can contribute to guaranteeing ethical behaviours in AI systems, even though it's not flawless. #tech #news #ai #generative #invention #practice #newton #rex #adobe #content #future #artist #company #tiktok #openai #research #music

Brave Search now provides code-related answers. A brand-new AI function called CodeLLM has been added to Brave Search to offer superior responses to questions about coding. CodeLLM creates step-by-step explanations for common coding challenges and highlights possible answers found online. It attempts to reduce the amount of time developers must spend sorting through numerous search results by providing succinct and useful answers. A sizable language model known as Mixtral, which can produce code from text inputs, powers the AI. By offering real-time context from its index of more than 20 billion online pages, Brave Search expands upon Mixtral. Brave Search offers CodeLLM for desktop and mobile users. Users only need to conduct a standard search using either the address bar, if Brave is set as the default search engine, or by going to search.brave.com to activate it. When a query qualifies for CodeLLM, Brave Search will recognise this automatically and show an answer box above the standard results. There is a button at the bottom of the widget where users can offer feedback on the functionality. CodeLLM advances Brave Search’s objective to develop in ways that respect user privacy and minimise unnecessary tracking. Without transferring any personal data to Brave or outside services, the feature operates only on the smartphone. Recent innovations that adhere to the same privacy-first policy include an AI summarizer that offers succinct summaries of search results and Goggles, which display alternative search rankings determined by the community. The introduction of CodeLLM emphasises how much better it is for Brave Search to control its whole technology stack instead of depending on outside vendors. It enables quick experimentation to produce unique search experiences that aren't available on market leaders like Google and Bing. Today, Brave Search processes more than 25 million requests per day from its base of more than 60 million Brave browser users, for whom it is the default setting. CodeLLM is another example of Brave Search's dedication to user-respecting innovation by directly empowering developers with AI power when they need it most, all without sacrificing their privacy. #tech #news #brave #search #engine #code #llm #large #language #model #user #ai #power #privacy #development #

The company states that the SEC's account compromise was "not due to any breach of X's systems." Late on Tuesday, social media platform Social Media X announced that it had finished its initial investigation into the U.S. Securities and Exchange Commission's hacked account, which featured a bogus post stating the SEC had authorised bitcoin exchange-traded funds for trading. X confirmed that the SEC's account had been compromised in the post, saying, "Based on our investigation, the compromise was not due to any breach of X's systems, but rather due to an unidentified individual obtaining control over a phone number associated with the @SECGov account through a third party." "We can also confirm that at the time the account was compromised, two-factor authentication was not enabled," X wrote in the post. After the unauthorised publication, the price of bitcoin surged, but it quickly dropped below $46,000 when the SEC stated that the bitcoin ETF was still pending approval. It was trading at approximately 12:20 a.m. ET, close to $45,958. The @SECGov X/Twitter account of the SEC has been hacked. The SEC or any of its employees did not make the unauthorised tweet about bitcoin ETFs, an SEC representative informed CNBC on Tuesday afternoon. SEC Chair Gary Gensler refuted the fake social media post claiming the commission has authorised bitcoin ETFs for trade. The regulator's approval of the bitcoin ETF is anticipated by the market. After years of opposition, the SEC is set to reach a judgement this week. Under Gensler's leadership, the SEC has pursued legal action against cryptocurrency exchanges like Coinbase and Ripple, alleging that they are selling unregistered securities. #tech #news #update #cyber #security #sec #x #company #legal #trade #authority #system #bitcoin #ripple #leader

TCS and AWS collaborated to establish a new practice focused on generative AI. In order to assist clients in realising the full potential of artificial intelligence (AI) and AWS generative AI services to revolutionise various aspects of their value chain and produce better business outcomes, Tata Consultancy Services (TCS) has established its AWS generative AI practice. Businesses all across the world are paying attention to generative AI because of its potential to drastically change industries. TCS has been leading the way in assisting clients from a variety of industries to investigate pertinent use cases for generative AI via pilots and proofs of concept. TCS has created a vast list of generative AI use cases by leveraging its in-depth domain expertise across several industry verticals. TCS has committed to providing over 100,000 workers with generative AI foundation training in order to expedite its customers' journeys. With the launch of this new practice today, as well as the certification of over 25,000 personnel in AWS generative AI services, their current focus is on significantly expanding their knowledge.  Using AWS technologies like Amazon Bedrock, TCS's generative AI practice for AWS will assist businesses in selecting and swiftly scaling the best solutions for their particular business needs and transforming their operations. TCS consultants will work with customers to co-innovate generative AI-powered solutions, experiment cooperatively, and investigate the most effective use cases in their business context. TCS Pace PortsTM, the company's innovation and research hubs situated in major city hubs like as New York, Pittsburgh, Toronto, Amsterdam, and Tokyo, are a suitable venue for co-innovation. The teams can also benefit from the work being done at these locations by university researchers and startup partners from TCS's innovation ecosystem. In addition to consulting and advice, big language model training and fine-tuning, solution design and prototyping, guardrail agent design, project delivery, and continuous maintenance, TCS provides an extensive spectrum of generative AI services and solutions. To assist businesses in navigating the moral and responsible applications of AI, TCS is developing a responsible AI framework. Using that technology, TCS' AWS generative AI practice will assist clients in finding and categorising organisational information and extracting insights to improve their business decision-making or produce content. Employee productivity and customer experience are greatly improved by the resulting solutions. #tech #news #awt #tcs #collaboration #ai #practice #assist #client #technology #framework #information #update #insight

AI-Powered Tools From Ayraa Are Revolutionising Workspace Knowledge With a range of solutions aimed at improving workplace communication and information management, Ayraa distinguishes out in the AI space. It revolutionises the way information is accessible, shared, and displayed by integrating smoothly with platforms like Slack and providing cutting-edge tools like Search & Assist, Scribe, and Deck. Ayraa is revolutionising offices globally by prioritising security and user-friendly design, resulting in more connected and efficient workspaces. The AI Revolution in the Workplace with Ayraa AI's incorporation into workspaces ushers in a new era of efficiency and productivity. Leading this change is Ayraa, which provides a range of AI-driven solutions to improve communication and information management in contemporary workplaces. Ayraa's AI-driven solutions, as opposed to conventional approaches, mark a dramatic advancement in workplace technology. Dissecting Ayraa: Beyond Just Another AI Instrument Ayraa stands out in the crowded AI market thanks to its distinctive solution for workplace issues. It provides a variety of products, each geared towards particular facets of communication and knowledge management. These tools, which each cater to different demands in the contemporary workstation, are Search & Assist, Scribe, and Deck. Discover & Help: Obtaining Solutions at the Speed of Thought The Search & Assist function of Ayraa completely reimagines how information is obtained and used in the workplace. With this tool, users can: Acquire pertinent data quickly from large data repositories. Make judgements based on accurate and current information. Cut down on the time you spend looking up information to increase productivity. Search & Assist's accuracy and speed revolutionise information retrieval, making it an indispensable tool in hectic work settings. Scribe: The AI-Driven Note Taker Another cutting-edge product from Ayraa, Scribe, simplifies the writing and delivery of strategic memoranda. Among its abilities are: jotting down important points from talks and meetings. arranging observations and insights into a logical, understandable structure. assisting teams and departments in exchanging knowledge. Scribe has a significant impact on communication and knowledge sharing by making sure that important information is effectively recorded and shared. #tech #news #update #upgrade #ayraa #search #engine #result #ai #structure #important #efficiency #ability #time #advancement

This Development May Significantly Affect Computer Vision and Other Fields A research team of worldwide scientists lead by Professor David Moss of Swinburne University of Technology has created an ultra-fast photonic signal processor that is 10,000 times faster than a traditional electronic signal processor, shattering the previous record. It can process nearly 400,000 videos at a rate of 17 terabits per second simultaneously. The utilisation of an integrated Kerr soliton crystal microcomb source in the device, which produces 95 different light wavelengths, is the team's primary innovation. This allows massively parallel computing in addition to the ultra-fast switching and transmission speeds of optics for the device. AI, machine learning, and computer vision will be significantly impacted by the speed and efficiency of the CPU. Large-scale real-time data processing will enable the development of more intelligent, self-governing robotic systems that perform well in practical settings. For instance, in astronomy, analysing massive amounts of cosmic data at previously unheard-of speeds would enable ground-breaking discoveries. This strategy will help speed up the operation of driverless cars and medical imaging. #tech #news #updates #ai #artifical #intelligence #computer #vision #data #processing #astronomy #cosmic #government #discovery #research #development

Microsoft and OpenAI are being sued by The New York Times for copyright violations. The artificial intelligence technology used by OpenAI and Microsoft to train ChatGPT and other services that give users instant access to information has been accused by the New York Times of copyright infringement. The technology is now in competition with the Times. The lawsuit was filed because the companies claim that their technology illegally copied millions of Times articles. The lawsuit is the most recent in a series that aim to restrict the use of purportedly extensive information scraping from the internet — without payment — to train so-called large language artificial intelligence models. Actors, authors, journalists, and other creative professionals who publish their works online worry that artificial intelligence (AI) would use their work as a basis and create rival chatbots and other information sources without paying them fairly. However, the Times is the first of the big news organisations to file a lawsuit against Microsoft and OpenAI, the two most well-known AI companies. Microsoft (MSFT) is a multibillionaire investor in OpenAI and holds a seat on the board of the firm. The Times stated in a case it filed on Wednesday that although it has an obligation to notify its subscribers, its capacity to do so is threatened by Microsoft and OpenAI's "illegal use of The Times's work to create artificial intelligence products that compete with it." Although OpenAI and Microsoft "given Times content particular emphasis" in their "widespread copying," the publication pointed out that they also used other sources in their attempt "to free-ride on The Times’s massive investment in its journalism by using it to build substitutive products without permission or payment." OpenAI said in a statement through spokesperson Lindsey Held, "We respect the rights of content creators and owners and are committed to working with them to ensure they benefit from AI technology and new revenue models." We've been having fruitful and forward-looking conversations with the New York Times, so we're shocked and dismayed by this development. As we do with many other publishers, we hope that we can work together in a way that benefits both of us. Requests for comment regarding the lawsuit were not answered by Microsoft. The Times said in its complaint that it was offended to learn months earlier that its work had been utilised to train the massive language models of the firms. The Times claimed that it started talks with Microsoft and OpenAI in April in order to establish conditions of an agreement and obtain just compensation. #tech #news #update #law #microsoft #openai #ai #copyright #request #cnn #times #new #york #talk #llm #models