Jul 17, 2024

Microsoft’s AI-Powered Designer Is Out of Preview And Available on iOS

Microsoft's AI-powered Designer app launches on iOS and Android, enabling users to create and edit images, designs, and more on the go. The app, which has been in preview for nearly a year, is now generally available with a range of AI-driven features. Key features and capabilities: Designer offers a comprehensive set of tools for creating and editing images and designs: Users can generate custom images, stickers, greeting cards, invitations, and more using AI-powered templates. The app includes an avatar creator, allowing users to design personalized avatars. Designer enables AI-driven image editing, such as restyling existing images, removing backgrounds, erasing...

read Jul 17, 2024

How SAP Is Using Gen AI To Transform Business Across Its Portfolio

SAP Labs is harnessing the power of generative AI to enhance business processes across industries, from e-sports to guitar manufacturing, with the hopes of driving unprecedented innovation and efficiency. Scaling AI across the enterprise: SAP Labs is focused on quickly integrating generative AI across its broad software portfolio, with insights from one customer's AI adoption being directly transferable to others: Over 24,000 SAP customers are currently using the company's AI solutions, with AI integrated into more than 10,000 published and 5,000 scenario-based interfaces and messages. SAP's AI in Concur Invoice Management automated the processing of 40 million invoices in 2022,...

read Jul 17, 2024

OpenAI Founding Member Announces Launch of New AI Education Startup Eureka Labs

Andrej Karpathy, former OpenAI founding member and Tesla AI scientist, has announced the launch of Eureka Labs, an AI-native school that aims to improve education by combining human expertise with AI teaching assistants. Teacher + AI symbiosis: Eureka Labs seeks to provide an ideal learning experience by leveraging AI to scale and guide human expert-written course materials: Subject matter experts will design course materials, while AI teaching assistants will guide students through them, allowing for a common platform to run an entire curriculum. The goal is to make it easy for anyone to learn anything, expanding education in both reach...

read Jul 17, 2024

A16Z, Silicon Valley Billionaires Are Aligning with Trump, Signaling Shift in Tech Industry Politics

Silicon Valley's billionaires are increasingly aligning with Donald Trump as he selects a former venture capitalist as his running mate, signaling a shift in the tech industry's political landscape. Key players supporting Trump: Prominent venture capitalists Marc Andreessen and Ben Horowitz have announced their intention to donate to Trump's election campaign, following Elon Musk's endorsement and plans to form a pro-Trump PAC: Andreessen and Horowitz, known as "techno-optimists," believe that the Biden administration's policies on AI and crypto have favored incumbents over startups, making Trump the "right choice" for what they call "little tech." The pair's "Little Tech Agenda" criticizes...

read Jul 17, 2024

FCC Proposes AI Disclosure Rule for Robocalls, Aiming to Protect Consumers

The FCC chair has proposed a new rule requiring robocalls to disclose the use of artificial intelligence (AI), aiming to protect consumers and enable informed decisions regarding these automated calls. Key elements of the proposed rule: Callers would need to obtain prior express consent to disclose their intent to use AI-generated calls. On each call, callers would be required to disclose to consumers when they are receiving an AI-generated call. The rule would provide a definition for AI-generated calls as the FCC seeks to establish guardrails around the use of this emerging technology in robocalls. Rationale behind the proposal: FCC...

read Jul 17, 2024

Salesforce Debuts AI-Powered Einstein Service Agent

Salesforce introduces Einstein Service Agent, a generative AI-powered self-service platform that empowers customers to resolve issues through conversational AI, marking a significant step forward in customer service automation. Key features of Einstein Service Agent: The new AI agent leverages Salesforce's extensive customer data and approved workflows to provide personalized, multi-step interactions for end-customers: Enables customers to ask questions, get resolutions, and take actions such as product returns or refunds through a conversational AI interface Seamlessly integrates with Salesforce's existing customer data and workflows, ensuring secure and trusted interactions grounded in the company's established systems Initially focused on customer service scenarios,...

read Jul 17, 2024

AI in Law: Navigating the Ethical Minefield of Automation Bias

The rapid advancements in AI and automation pose significant ethical challenges, particularly in the legal sector, where automation bias and complacency can have far-reaching consequences. The legal industry's vulnerability to AI-related issues: As a traditionally slow adopter of technology, the legal sector is now grappling with the profound impact of AI on knowledge-based tasks: Lawyers, often considered the "canaries in the coal mine" for knowledge work, are highly susceptible to machines providing answers more quickly, making automation bias and complacency substantial threats to the profession. The assumption that younger, digital-native generations are better equipped to discern fake information and guard...

read Jul 17, 2024

Cohere and Fujitsu Partner to Bring Localized AI Solutions to Japanese Enterprises

Cohere, a Canadian AI startup, has partnered with Japanese tech giant Fujitsu to develop large language models (LLMs) and AI solutions tailored for the Japanese market, aiming to empower local enterprises with advanced natural language processing capabilities. Key details of the partnership: Cohere will provide its cutting-edge AI models, while Fujitsu will leverage its expertise in Japanese language training and fine-tuning technologies to create enterprise-ready solutions: The collaboration will debut an LLM tentatively named 'Takane' (meaning 'mountain peak' in English), based on Cohere's enterprise-grade Command R+ model, which offers best-in-class retrieval-augmented generation (RAG), multilingual coverage, and a powerful Tool Use...

read Jul 17, 2024

FTC Probes Amazon’s Deal With Adept Amid Heightened Scrutiny of Tech Giants

The FTC is scrutinizing Amazon's deal with AI startup Adept, seeking more information about the tech giant's move to hire most of Adept's team and license its technology, according to an exclusive Reuters report. Key details of the deal: Amazon's acquisition playbook involved hiring Adept's team and licensing its AI technology: Adept announced that its plans to develop "useful general intelligence and an enterprise agent product" would have required significant focus on fundraising, leading to the deal with Amazon. The FTC's informal inquiry may not necessarily lead to a full investigation or enforcement action, but it highlights the close attention...

read Jul 17, 2024

Tinder’s AI Can Now Choose Your Most Flattering Photos to Boost Matches

Tinder's AI Photo Selector can now choose your most flattering profile pictures by analyzing your entire camera roll, aiming to help users put their best face forward and potentially increase their matches. How it works: The AI feature requires a sample photo of your face, then scans your phone's photos to find the best ones of you based on factors like lighting and composition: Users can either let the AI analyze their entire camera roll or select a smaller batch of photos for it to choose from. The AI has been trained to avoid picking photos that violate Tinder's rules,...

read Jul 17, 2024

French AI Startup Mistral Launches Groundbreaking Open-Source Models for Coding and Math

Mistral, a well-funded French AI startup, has launched two new large language models (LLMs) based on the Mamba architecture, which aims to improve upon the efficiency of transformer-based models. Codestral Mamba for developers: Mistral's code-generating model, Codestral Mamba 7B, offers faster response times and longer input handling compared to other models: The model can handle inputs of up to 256,000 tokens, double that of OpenAI's GPT-4. In benchmarking tests, Codestral Mamba outperformed rival open-source models like CodeLlama 7B, CodeGemma-1.17B, and DeepSeek in HumanEval tests. Developers can modify and deploy the model from its GitHub repository and through HuggingFace, with an...

read Jul 17, 2024

Microsoft’s AI-Powered Data Governance Solution Sees Rapid Adoption Amid Enterprise AI Push

Microsoft's Purview Data Governance solution sees rapid adoption amid enterprise demand for robust AI data management tools. Federated governance model balances business autonomy and centralized control: Microsoft's approach allows different business units to manage their own data products while maintaining centralized policy enforcement: The federated model aims to empower business users who best understand how to create their data products, while still ensuring organizational control. Purview uses natural language processing to enhance data visibility and management, allowing users to interact with governance tools using natural language queries, potentially lowering barriers to adoption. AI is being leveraged at nearly every layer...

read Jul 16, 2024

Hacker Group Leaks Disney Data, Igniting More AI Debate in Hollywood

Hacker group NullBulge claims responsibility for Disney data leak: NullBulge, a self-proclaimed hacktivist group, says it breached thousands of Disney's internal messaging channels and leaked approximately 1.2 terabytes of data, including computer code, information about unreleased projects, and conversations about marketing, studio technology, and job applicants. Leak motivated by concerns over Disney's AI practices and treatment of artists: According to NullBulge, the leak was prompted by Disney's handling of "artist contracts, its approach to AI, and its pretty blatant disregard for the consumer": The group, which claims to be based out of Russia, says it gained access to Disney's system...

read Jul 16, 2024

Huma Raises $80M To Accelerate Development of Digital Health Apps

Huma, a London-based health-tech company, has launched an AI-powered cloud platform that enables startups to quickly prototype, launch, and scale regulated digital health apps, including those enhanced with AI capabilities. Generative AI builder experience drives rapid app development: The Huma Cloud Platform offers a suite of tools, most notably a generative AI-powered builder that can transform text prompts into functional applications within minutes: Users can create custom digital health apps using a combination of widgets, modules, templates, and disease management tools through a simple drag-and-drop interface. If drag-and-drop functionality isn't sufficient, users can leverage the generative AI builder by inputting...

read Jul 16, 2024

Microsoft’s AI Hiring Spree Sparks UK Antitrust Probe

The UK's Competition and Markets Authority (CMA) has formally launched an investigation into Microsoft's hiring of executives from AI startup Inflection to determine if the move could undermine competition in the UK market: In March, Microsoft hired two Inflection AI co-founders, Mustafa Suleyman and Karén Simonyan, to lead its new Microsoft AI division, along with several other Inflection staff members. The CMA's preliminary investigation, started in April, aimed to assess whether these hires and Microsoft's partnership with French AI startup Mistral could shield the tech giant from competition. The formal "merger inquiry" will conclude by September 11, when the CMA...

read Jul 16, 2024

Survey Finds Most People Believe AI Chatbots Are Already Conscious

A recent survey conducted by the University of Waterloo found that most people believe generative AI chatbots like ChatGPT are conscious, despite expert consensus to the contrary, highlighting AI's remarkable ability to mimic human-like interactions. Key findings: Two-thirds of participants believe AI is conscious; The study, which surveyed 300 people, revealed that 67% believe ChatGPT and other AI chatbots can reason, feel, and be aware of their own existence in some form. This perception of AI consciousness was more prevalent among frequent users of AI tools, demonstrating the convincing nature of ChatGPT's human-like responses. However, experts emphasize that current AI...

read Jul 16, 2024

Claude Challenges ChatGPT on Android In Anthropic’s Mobile AI Gambit

Anthropic has released its AI assistant Claude on Android, intensifying competition with OpenAI's ChatGPT and signaling a new phase in the mobile AI race. Mobile AI showdown: Claude's Android debut targets ChatGPT's dominance; The Android launch is a strategic move by Anthropic to rapidly scale its user base and challenge OpenAI's market leadership: With over 3 billion active Android users worldwide, Anthropic aims to close the gap with ChatGPT in terms of mobile accessibility and adoption. Anthropic's strict privacy policy of not using user interactions for model training could appeal to privacy-conscious users and enterprises, differentiating Claude in the competitive...

read Jul 16, 2024

Haiper 1.5 Is the Latest Challenger in the Text-to-Video Battleground

The launch of Haiper 1.5 positions the London-based startup as a strong contender in the rapidly evolving AI video generation market, taking on established players like Runway, Pika, and OpenAI: Haiper, founded by former Google Deepmind researchers, has onboarded over 1.5 million users on its platform just four months after emerging from stealth, highlighting its strong positioning and potential for growth. Runway recently opened access to its new, more realistic model for video generation, setting a high bar for competitors in terms of quality and capabilities. OpenAI's entry into the text-to-video space with its Sora model capable of generating 60-second...

read Jul 16, 2024

Arcee AI Secures $24M, Signaling Investor Confidence in Small Language Models

Arcee AI, a startup specializing in small language models (SLMs), has secured $24 million in Series A funding, signaling growing investor confidence in the potential of more efficient and specialized AI models for enterprise applications. Arcee AI's unique approach: The company focuses on developing domain-specific SLMs and providing tools for enterprises to create their own customized models, enabling organizations to tackle multiple use cases cost-effectively: Arcee AI's Model Merging technology allows the combination of multiple AI models into a single, more capable model without increasing its size, creating highly tailored models that can outperform larger, more general models in specific...

read Jul 16, 2024

NVIDIA AI Supercharges Filmora Video Editing, Enabling HDR and Accelerated Workflows

The video editing software Wondershare Filmora has added support for NVIDIA's AI-powered RTX Video HDR technology, enabling users to significantly enhance the quality and vibrancy of their videos. Key features and updates: Wondershare Filmora now supports NVIDIA RTX Video HDR, which uses AI to transform standard dynamic range video into HDR10-quality video with expanded color gamut, clearer frames and greater depth. Filmora users with NVIDIA RTX GPUs can also leverage RTX Video Super Resolution to further sharpen and enhance video quality by restoring features and removing artifacts. Twitch Enhanced Broadcasting has been added to OBS Studio and XSplit Broadcaster, giving...

read Jul 16, 2024

Game Developers Fear AI Will Lead To Job Losses and Reduced Creativity

The AI boom in gaming is causing concern among game developers about the future of their jobs and the industry as a whole: Key takeaways: While some see AI as a potential cost-saving solution for the gaming industry, many developers worry it could lead to job losses, reduced creativity, and a focus on fixing AI-generated content rather than creating original work: Game artist Jess Hyland says she's "very aware" her job could disappear, and that "everyone knows someone who's been laid off" amid widespread industry layoffs and studio closures. There are accounts suggesting entry-level roles like concept art have already...

read Jul 16, 2024

New Research Highlights How “FlashAttention-3” May Make Training and Inference More Efficient

FlashAttention-3, a new technique developed by researchers from multiple institutions, dramatically accelerates attention computation on Nvidia's H100 and H800 GPUs, enabling faster and more efficient training and inference of large language models (LLMs). The challenge of attention computation in LLMs: As LLMs grow larger and process longer input sequences, the computational cost of the attention mechanism becomes a significant bottleneck due to its quadratic growth with sequence length and reliance on operations not optimized for GPUs. Attention computations involve a mix of matrix multiplications and special functions like softmax, which are computationally expensive and can slow down the overall computation...

read Jul 16, 2024

Microsoft’s New SpreadsheetLLM Offers Glimpse Into Future of Data Interaction

Microsoft researchers propose SpreadsheetLLM, a novel method that helps AI models understand and process spreadsheets more efficiently, potentially improving chatbot interactions with complex data. Key innovation: SheetCompressor framework: Microsoft's SheetCompressor encoding framework compresses spreadsheets into bite-sized chunks that large language models (LLMs) can more easily handle: It includes modules that make spreadsheets more legible for LLMs, bypass empty cells and repeating numbers, and help LLMs better understand the context of numbers (e.g., distinguishing years from phone numbers). This compression method reduced token usage for spreadsheet encoding by up to 96%, significantly boosting performance on larger spreadsheets where high token usage...

read Jul 16, 2024

Meta’s Upcoming LLM Launch Will Be Another Big Milestone for Open-Source AI

Meta is set to release Llama 3 400B, its most powerful open-source AI language model, by the end of July 2024. Key details and capabilities: Meta's upcoming release of the Llama 3 400B model is highly anticipated due to its impressive performance and open availability for research and commercial use: Boasting over 400 billion parameters, Llama 3 400B achieves near-parity with OpenAI's GPT-4 on the MMLU benchmark despite using less than half the parameters, suggesting significant advancements in model architecture and training efficiency. The model promises new capabilities such as multimodality, multilingual conversation, longer context windows, and stronger overall performance...

read