This week's latest AI industry updates - December 3, 2024

Stay informed with SymphonyAI's weekly Eureka newsletter, covering all the important AI industry developments and technology advancements you need to know.

SymphonyAI news?

• Financial Services—Customer news: Bank of Bahrain and Kuwait (BBK) is implementing the SymphonyAI AI-based anti-money laundering?transaction monitoring and?customer due diligence?technology?

• Enterprise IT—Coverage: 11 ways to ensure IT-business alignment?

• Enterprise IT—Upcoming webinar: Gen AI Capability Adoption?

• Industrial—Upcoming webinar: Roadmap to Smart Maintenance Planning

Big tech announcements?

• Microsoft’s research points to the future of GUI automation:?Microsoft researchers report that AI agents can now control computer interfaces and execute complex tasks via natural language commands.?

• OpenAI moves to trademark its o1 ‘reasoning’ models: OpenAI has filed a trademark application for "OpenAI o1," its new reasoning AI model, as part of its ongoing efforts to protect its intellectual property.?

• OpenAI faces critical test as Chinese models close the gap in AI leadership: OpenAI faces increasing competition from rapidly advancing Chinese AI models and open-source initiatives, putting pressure on the company to maintain its leadership position in the field of advanced AI reasoning capabilities. ?

• Amazon adds automatic prompt optimization to its Bedrock AI service: Amazon has introduced automatic prompt optimization for its Bedrock AI service, allowing users to enhance AI model performance across various tasks with a single API call or button click, simplifying the prompt engineering process.?

• Amazon Connect Puts Generative AI to Work Improving End-to-End Customer Experiences: Amazon Connect introduces new generative AI features, including proactive customer outreach, improved self-service, AI safeguards, and AI-powered agent evaluation tools, to enhance customer service and operational efficiency for contact centers.?

• Meta plans to build a $10B subsea cable spanning the world, sources say: Reports suggest Meta plans to build a $10+ billion, 40,000+ kilometer global subsea fiber-optic cable, marking its first solely-owned major infrastructure project to support its growing internet traffic and AI ambitions.?

• Uber branches out into AI data labeling:?Uber is expanding its gig economy model by creating a new division called Scaled Solutions, which hires contractors for AI annotation and data labeling tasks for both internal projects and external clients.

AI impact?

• AI Infrastructure Spending to Surpass $100B by 2028: According to an IDC report, global spending on AI infrastructure is projected to exceed $100 billion by 2028, driven by significant growth in server and storage investments for AI deployments across cloud and enterprise environments.?

• Cisco Reports Decline in AI Readiness Among Canadian Organizations: Cisco's 2024 AI Readiness Index found that only 7% of Canadian organizations are fully prepared to deploy AI technologies, down from 9% last year, highlighting a growing gap between AI adoption intentions and actual implementation across sectors.?

• CIOs contend with gen AI growing pains: Among a lack of a clear framework for generative AI success, IT leaders are beginning to identify key components of a gen AI roadmap—including data, platform, and skills—while grappling with the challenges of moving from proof of concept to production.?

AI agents: The next wave of AI?

• 2025—The year ‘invisible’ AI agents will integrate into enterprise hierarchies: In 2025, AI agents are expected to become more visible and integrated into enterprise operations, performing complex tasks and collaborating with human workers across various business functions.?

Noteworthy AI applications?and use cases

• World Labs’ AI can generate interactive 3D scenes from a single photo: World Labs, founded by AI pioneer Fei-Fei Li, has introduced an AI system that can generate interactive, modifiable 3D scenes from a single image, aiming to revolutionize the creation of virtual environments.?

• ConcertAI Expands its AI-Enhanced Clinical Capabilities in Medical Imaging Machine Learning: ConcertAI's TeraRecon launches cloud-based AI imaging solutions, offering flexible deployment and FDA-cleared cardiac MR auto-contouring, enhancing accessibility and capabilities for healthcare providers.?

Platform and model developments?

• Anthropic announces custom styles for Claude.ai users: Anthropic has introduced custom styles for Claude.ai, allowing users to personalize the AI's responses to match their preferred communication style, tone, and structure for various workflows and contexts. ?

• Anthropic releases open-source Model Context Protocol: Anthropic has open-sourced the Model Context Protocol (MCP), a universal standard that allows AI assistants to securely connect with various data sources, aiming to improve AI responses by providing better access to relevant information.??

• Alibaba releases an open reasoning model that beats o1-preview: Alibaba has released Qwen with Questions (QwQ), an open-source large reasoning model that outperforms OpenAI's o1-preview on certain mathematical and scientific reasoning benchmarks. ?

• Alibaba researchers unveil Marco-o1, an LLM with advanced reasoning capabilities: Marco-o1 enhances reasoning capabilities for complex, open-ended problems using advanced techniques like Monte Carlo Tree Search and self-reflection mechanisms.?

• Ai2 releases new language models competitive with Meta’s Llama: Ai2 has released OLMo 2, a new family of open-source AI language models that can be fully reproduced from scratch, meeting the Open Source Initiative's definition of open-source AI.?

• Hugging Face releases SmolVLM: SmolVLM is a compact and efficient vision-language AI model that processes images and text with high performance while requiring significantly less computational power than larger competitors, potentially cutting AI costs for businesses.

• Perplexity Leverages Quartr API to Power First-of-its-Kind Analysis Tools for Retail Investors: Quartr, a provider of qualitative research tools, has partnered with Perplexity to offer live earnings call transcripts and AI-generated summaries to Perplexity's users, enhancing access to real-time corporate insights.?

• Linkup connects LLMs with premium content sources (legally): Linkup is developing an API that allows AI developers to legally access premium content from trusted sources to enhance their large language models' outputs, acting as a marketplace between content publishers and AI companies.?

AI regulations and public policy?

• US cuts off China's access to chip manufacturing equipment in bid to limit AI capabilities: The US Department of Commerce has imposed its most extensive restrictions on semiconductor exports to China, targeting chip manufacturing equipment, software, and companies to limit China's ability to produce advanced AI chips domestically.?

• Elon Musk files for injunction to halt OpenAI’s transition to a for-profit: Elon Musk's attorneys have filed for a preliminary injunction against OpenAI, its co-founders, and Microsoft, seeking to halt alleged anticompetitive behavior and preserve OpenAI's nonprofit status.?

• Canadian media companies sue OpenAI over copyright claims: Five major Canadian media organizations have filed a lawsuit against OpenAI, alleging that the company used their copyrighted content without permission to train ChatGPT.?

Research ?

• ChatGPT search provides inaccurate sources even for OpenAI partner publishers: A Columbia University study found that ChatGPT's search function frequently provides inaccurate information and fabricates sources, even for content from publishers with OpenAI agreements, raising concerns about its reliability for news citations.?

• LLMs can outperform neuroscientists at predicting research outcomes: A University College London study found that large language models outperform human experts in predicting scientific research outcomes, achieving 81.4% accuracy compared to 63.4% for experienced neuroscientists across various experimental scenarios.?

• Adobe's "MultiFoley" AI creates synchronized sound effects for video: Adobe Research and University of Michigan have developed MultiFoley, an AI system that generates high-quality, synchronized Foley sound effects for videos using text prompts, reference audio, or video examples.?

Funding?

• European AI infra company Nebius nabs $700M from Nvidia, Accel, others: Nebius, formerly known as Yandex N.V., has raised $700 million from investors including Nvidia, Accel, and Orbis to fund its expansion into the U.S. AI infrastructure market. ?

• AI agent startup /dev/agents commands a $56M seed round at a $500M valuation: Former Google executives have launched /dev/agents, raising $56 million to develop a new operating system for AI agents, aiming to provide a unified platform for developers to build and integrate AI programs that can collaborate on complex tasks. ?

• PlayAI Secures $21 Million in Seed Funding: PlayAI, a conversational voice AI startup, raises $21 million in seed funding to develop real-time voice agents and interfaces for businesses, developers, and creators, while also releasing a new context-aware speech model for more natural conversations.?

• AI SPERA Secures $9 Million Investment: South Korean AI cybersecurity company AI SPERA raises $9 million in funding, impressing investors with its global success as 90% of its users come from 150 countries worldwide. ?