Gen AI for Business #8

Explore the latest in generative AI for business! We’re bringing you the freshest curated news, smart strategies, and groundbreaking innovations shaking up the B2B sector. Whether it’s cutting-edge tech developments or practical tools to boost your business, we've got you covered.

From energy-saving AI models to game-changing applications in healthcare and legal sectors, there's a lot to unpack. Whether you’re here for the latest trends, curious about AI in your industry, or just love staying ahead of the curve, we’ve got something for you.?

If you enjoyed this letter, please leave a like or a comment and share! Knowledge is power.?

Thank you for reading,

Eugina

News about models and everything related to them

Discover the latest in AI with this exciting roundup! IEEE Spectrum unveils 1-bit LLMs slashing energy use without losing performance. MarkTechPost lists 12 hot LLM leaderboards, and DataScienceCentral explores modular, multi-modal AI models. JIVI’s LLM outshines GPT-4 in medical Q&A, and researchers warn of "LLMjacking" schemes costing up to $46K daily. The HippoRAG framework, inspired by neurobiology, boosts LLM memory, while Qwen2's multilingual models excel in coding and long-context tasks.

• 1-bit LLMs Could Solve AI’s Energy Demands - IEEE Spectrum? A recent article from IEEE Spectrum discusses the potential of 1-bit large language models (LLMs) to address AI's energy demands. By reducing the precision of calculations to 1-bit, these models can significantly lower power consumption and computational requirements while maintaining performance. This approach could make AI more sustainable and cost-effective, which is crucial as AI applications expand. The development of 1-bit LLMs marks a significant advancement in creating more efficient AI systems.

• Top 12 Trending LLM Leaderboards: A Guide to Leading AI Models' Evaluation - MarkTechPost? 12 trending leaderboards for evaluating large language models (LLMs), including the Open LLM Leaderboard, MTEB, HELM, Chatbot Arena, Chatbot Battle, CodeEval, CodexEval, Function Calling, HumanEval, LM Evaluation Harness, MT-Bench, and RAG-Leaderboard. These leaderboards assess models on various tasks such as text embedding, code generation, and function-calling capabilities, providing comprehensive evaluations to distinguish true advancements in AI.

• New Trends in LLM Architecture - DataScienceCentral.com? Recent trends in large language model (LLM) architecture highlight several key innovations. Modular design is becoming prominent, allowing for specialized components within the models. Multi-modal capabilities are being integrated, enabling models to process text, images, and other data types simultaneously. Efficiency improvements, such as sparsity and quantization, are being pursued to reduce computational demands. Ethical considerations and transparency are increasingly prioritized in model development. These trends aim to enhance performance, versatility, and responsible AI usage.

• AI Startup Jivi’s LLM Beats OpenAI’s GPT-4 & Google’s Med-PaLM 2 in Answering Medical Questions - insideBIGDATA AI startup JIVI's large language model (LLM) outperformed OpenAI's GPT-4 and Google's Med-PaLM 2 in answering medical questions. This breakthrough demonstrates JIVI's advanced capabilities in providing accurate and reliable medical information, setting a new standard in the field of AI-driven healthcare solutions.?

• Researchers Uncover 'LLMjacking' Scheme Targeting Cloud-Hosted AI Models? Researchers have uncovered a "LLMjacking" scheme where attackers use stolen cloud credentials to access and monetize cloud-hosted large language models (LLMs). By exploiting vulnerabilities in the Laravel Framework and using AWS credentials, attackers target local LLM models, like Anthropic's Claude, without running legitimate queries. This can lead to significant financial losses for victims, with potential costs exceeding $46,000 per day. Organizations are advised to enhance logging, monitor for unauthorized activities, and strengthen vulnerability management to mitigate these risks.

• Neurobiological Inspiration for AI: The HippoRAG Framework for Long-Term LLM Memory - MarkTechPost? The HippoRAG framework, inspired by neurobiological principles, addresses long-term memory challenges in large language models (LLMs) by enhancing knowledge integration. Developed by researchers from Ohio State University and Stanford University, HippoRAG uses a graph-based hippocampal index to create associations between passages, improving retrieval and synthesis of information. It significantly outperforms existing methods in tasks like multi-hop question answering, providing a more efficient and cost-effective solution for integrating new knowledge without catastrophic forgetting.?

• Hello Qwen2 | Qwen? The Qwen team announced the release of Qwen2, featuring pretrained and instruction-tuned models across five sizes, supporting 27 languages beyond English and Chinese. These models exhibit state-of-the-art performance in coding, mathematics, and long-context understanding up to 128K tokens. Significant improvements in multilingual capabilities and reduced context-switching issues were achieved through enhanced training datasets. Qwen2 models outperform their predecessors and competitors in various benchmarks, ensuring better alignment with human values and safety.?

Gen AI news from different industries

Stanford's study warns of AI hallucinations in legal research, stressing accuracy checks. Early adopters in law streamline workflows but face privacy challenges. Generative AI revolutionizes insurance by tailoring policies and speeding up claims. MIT's AI framework maps new physical phases efficiently. Schools juggle AI for learning and integrity. Architects use AI for quicker designs, while media execs boost creativity with AI. Healthcare AI predicts patient outcomes, enhancing care.

Legal

• Hallucination-Free? Assessing the Reliability of Leading AI Legal Research Tools A Stanford University study found that generative AI tools used in the legal industry produce false or misleading information in 17% to 33% of cases.? AI legal tools frequently produce inaccurate or "hallucinated" answers, which can be problematic for legal professionals relying on these technologies. These hallucinations highlight the challenges and risks associated with using AI in legal contexts, stressing the need for careful oversight and validation of AI-generated content.

• Early Adopters Of Gen AI In Law? Early adopters of generative AI in law are leveraging the technology to enhance legal research, automate document review, and improve contract analysis. These implementations are streamlining workflows, reducing costs, and increasing efficiency. However, challenges such as ensuring data privacy and maintaining accuracy persist. The legal industry is cautiously optimistic, balancing innovation with the need for stringent ethical standards.?

Insurance

• How GenAI is redefining insurance - FinTech Global Generative AI is transforming the insurance industry by streamlining operations and enhancing customer interactions. It personalizes customer experiences by analyzing data to create precise risk profiles and tailored policies, boosting satisfaction and loyalty. In claims processing, AI reduces the lifecycle by synthesizing large volumes of data. It also aids in compliance monitoring, fraud detection, and creating training materials, thus mitigating risks. AI-driven automation frees up professionals for complex tasks, aligning with trends in digital customer experiences and advanced analytics, ensuring insurers remain efficient and customer-centric.

Physics

• Scientists use generative AI to answer complex questions in physics | MIT News | Massachusetts Institute of Technology MIT researchers have developed a generative AI-based framework to automatically classify phases in physical systems, outperforming traditional and existing machine-learning methods. This innovative approach, leveraging fewer data and integrating physical system knowledge, enables efficient mapping of phase diagrams and could advance studies in novel materials and quantum systems. The technique, more efficient and less reliant on large datasets, represents a significant step towards automated scientific discovery of new phases of matter.?

Education?

• AI and education: Schools grapple with challenging technology - GREENVILLE JOURNAL? Schools are grappling with the integration of AI technology like ChatGPT. While some institutions prohibit its use to prevent academic dishonesty, others are exploring supervised applications for educational enhancement. Educators are being trained to leverage AI for administrative tasks and personalized learning, despite concerns about plagiarism and ethical use. The overall goal is to balance the benefits of AI in education with maintaining academic integrity and teaching responsible usage.

Architecture

• Architects tap in to Gen AI for design and business efficiencies | RIBAJ Generative AI is transforming architecture by enhancing design and business efficiency. Architects use AI tools for quick idea generation, reducing time spent on inspiration and visualizations. AI helps in creating detailed early-stage visualizations, fine-tuning designs with tools like Stable Diffusion, and speeding up the production of architectural renders. Additionally, AI assists in administrative tasks, such as comparing tender returns and reviewing reports, allowing architects to focus on more complex activities. Despite its benefits, there are concerns about job displacement and maintaining design integrity.

Entertainment

• What Media & Entertainment Execs Are Saying About Using Gen AI: The VIP+/HarrisX Spring 2024 Survey Generative AI is significantly impacting the media and entertainment industry by enhancing content creation, improving efficiency, and offering personalized experiences. Executives report that AI tools are being used for generating creative content, such as scripts and marketing materials, and for automating mundane tasks like data analysis. These tools enable rapid idea generation and customization, making production processes faster and more cost-effective. AI is also being leveraged for personalized marketing, analyzing consumer behavior to create targeted campaigns that enhance engagement and satisfaction. Additionally, generative AI is facilitating the creation of realistic images and videos, transforming visual storytelling in film and media. However, concerns about job displacement and maintaining creative originality persist as the industry adapts to these technological advancements.

Healthcare

• https://academic.oup.com/jpubhealth/advance-article-abstract/doi/10.1093/pubmed/fdae078/7686453? Overview of the impact of AI on public health, focusing on its potential to enhance disease surveillance, healthcare delivery, and health research. It highlights AI's role in improving accuracy and efficiency in diagnosing diseases, predicting outbreaks, and personalizing treatments. However, the study also raises concerns about data privacy, ethical considerations, and the need for robust regulatory frameworks to ensure the safe and equitable implementation of AI in public health.?

• LORIS robustly predicts patient outcomes with immune checkpoint blockade therapy using common clinical, pathologic and genomic features | Nature Cancer? Researchers developed the LORIS score, a logistic regression-based tool, to predict patient outcomes with immune checkpoint blockade (ICB) therapy across 18 solid tumor types. LORIS, using six clinical, pathologic, and genomic features, outperforms existing models in identifying patients likely to respond to ICB, even those with low tumor mutational burden. It accurately predicts both short-term and long-term survival, facilitating precise patient stratification for ICB therapy, thus enhancing clinical decision-making in precision oncology.

• Key considerations to cancer institute’s gen AI deployment | CIO? Deploying generative AI in cancer institutes involves focusing on data governance, ensuring privacy and compliance, and building patient trust through transparency. Effective governance models and interdisciplinary collaboration are essential for scalable and ethical AI use. Partnerships with tech companies help integrate AI efficiently. AI applications, such as improving clinical documentation and enhancing diagnostic accuracy, reduce administrative burdens and improve patient care.?

Software Engineering

• Many developers still aren't really sure how useful Gen AI tools will be for them | TechRadar? A TechRadar survey reveals that while 76% of developers use or plan to use generative AI tools, many remain uncertain about their effectiveness. Approximately 38% are concerned about AI's accuracy, noting that AI can provide incorrect information at least half the time. Despite these concerns, 95% believe AI tools can enhance productivity, though measuring this productivity remains unclear. Popular tools include Codeium, GitHub Copilot, and ChatGPT, but issues with context and complexity persist.?

• McKinsey's article on generative AI (gen AI) highlights its transformative impact on the software industry. Gen AI is expected to drive significant growth, with enterprise spending potentially reaching $250 billion by 2027. Key impacts include accelerated vendor switching, value creation through new use cases, and easier software development. However, challenges include increased competition and potential value erosion as companies might shift from buying to building software. To thrive, software leaders need to rapidly adopt gen AI, reimagine product categories, reallocate resources, and innovate continuously. Navigating the generative AI disruption in software?

Product Management

• How generative AI could accelerate software product time to market Generative AI significantly accelerates software product time to market by enhancing the productivity and quality of product managers' work. McKinsey's study found that using generative AI tools improved PM productivity by 40%, shortened time to market by 5%, and enhanced the employee experience by 100%. These tools help automate high-toil tasks and enable PMs to focus on strategic activities, thereby improving overall product quality and customer satisfaction. Adopting a phased approach to integrating AI tools can maximize these benefits.?

Marketing

• Say Hello to My New AI Marketer: How Gen AI-Based Software Is Advancing Marketing and Sales | Andreessen Horowitz It highlights three phases of AI integration: using AI as marketing copilots for content creation and segmentation, evolving into marketing agents for hyper-personalized campaigns, and ultimately developing autonomous marketing teams. This evolution is driven by the need for scalable, personalized marketing strategies amid increasingly fragmented customer locations.?

Advertising

• Generative AI’s role in shaping the future of advertising? It allows for the development of AI assistants that provide continuous consumer engagement and memory capabilities, reducing the reliance on keywords. This technology facilitates hands-free interactions, enhances consumer memory, and decreases transactional friction, leading to higher conversion rates. Generative AI also promises to lower creative production costs, making custom ads more accessible. As a result, advertisers can deliver more relevant and effective campaigns, ultimately boosting engagement and loyalty.?

Manufacturing

• Harnessing AI to Streamline Manufacturing Processes? AI advancements, such as computer vision and natural language processing, enable precise quality assurance and better analysis of unstructured data. AI also enhances IoT analytics and knowledge management within factories. The key to leveraging AI's benefits lies in high-quality data collection and the continuous refinement of AI models to ensure accuracy and efficiency.?

Robotics

• The AI revolution is coming to robots: how will it change them?? AI advancements are transforming robotics, enabling robots to perform diverse, autonomous tasks. Researchers are embedding AI models with common-sense knowledge into robots, enhancing their versatility. Key challenges include the need for vast physical data, robust hardware, and ensuring safety. Despite these hurdles, AI-driven robotics hold the potential to create adaptable, general-purpose robots, marking a significant leap in the field.

?

Big News

Apple's integration of ChatGPT into iPhones with iOS 18 sparks privacy and financial concerns for Microsoft due to Apple's 30% in-app purchase fee. OpenAI revives its robotics research group, enhancing robot manipulation with generative AI. NVIDIA's NIM platform revolutionizes AI deployment, drastically cutting setup time, while their GeForce RTX AI PCs promise to transform gaming and content creation. AMD and Intel unveil new AI chips, aiming to challenge NVIDIA's dominance in the AI hardware market. Which of these advancements will impact the AI landscape the most?

• OpenAI's Apple Deal Sparks Concerns In Microsoft, Could Shift AI Market Dynamics? Apple is finalizing a deal with OpenAI to integrate ChatGPT into iPhones via iOS 18, aiming to enhance AI features on the platform. This partnership could position OpenAI's technology prominently on Apple devices, potentially debuting at the WWDC 2024. However, there are concerns about privacy, ethical implications, and the financial impact due to Apple's 30% in-app purchase fee.

• OpenAI is restarting its robotics research group? OpenAI is reviving its robotics research group after a three-year hiatus. This group will focus on applying generative AI, like ChatGPT, to robotics tasks, enhancing robot manipulation capabilities. They are hiring robotics engineers to develop multimodal models and improve core AI architectures. This move follows OpenAI's previous robotics efforts and aligns with their recent investments in the robotics sector, such as their participation in Figure AI's fundraising.

• NVIDIA NIM Revolutionizes Model Deployment, Now Available to Transform World’s Millions of Developers Into Generative AI Developers? NVIDIA has launched NIM (NVIDIA Inference Microservices) to streamline the deployment of generative AI applications. NIM allows developers to build and deploy models as optimized containers on various infrastructures, significantly reducing deployment time from weeks to minutes. The platform supports multiple AI modalities, enhancing productivity and efficiency. NIM's integration with major AI ecosystems and platforms, including AWS, Google Cloud, and Hugging Face, enables widespread adoption. Enterprises can leverage NIM for diverse applications, including text generation, digital avatars, and drug discovery.?

• NVIDIA Brings AI Assistants to Life With GeForce RTX AI PCs? NVIDIA introduced GeForce RTX AI PCs, featuring Project G-Assist and NVIDIA ACE NIM inference microservices, to enhance AI assistant capabilities and digital human interactions. Project G-Assist, demonstrated with ARK: Survival Ascended, provides context-aware gaming assistance. The RTX AI Toolkit and new AI-powered laptops from ASUS and MSI aim to optimize AI deployment and performance. This technology promises to transform gaming, content creation, and more by integrating advanced AI models and tools.?

• AMD unveils new AI chips to compete with Nvidia - Fast Company? AMD unveiled new AI processors at Computex, including the MI325X accelerator available in late 2024, aiming to compete with Nvidia's dominance in AI chips. AMD plans annual updates, highlighting AI as its top priority. Future chips, MI350 and MI400, promise significant performance improvements and new architectures, set for 2025 and 2026 releases, respectively. AMD expects $4 billion in AI chip sales for 2024, up from previous estimates. The company's new CPUs will also feature enhanced AI capabilities for PCs.?

• Intel unveils new AI chips as it seeks to reclaim market share from Nvidia and AMD? Intel has unveiled its new Gaudi 3 AI chips, designed to challenge Nvidia and AMD in the AI hardware market. The Gaudi 3 promises significant improvements in performance and energy efficiency, with Intel claiming it can be 1.5 to 1.7 times faster than Nvidia's H100 in training AI models. These chips are built on a five-nanometer process and feature enhanced networking capabilities, using 200 Gbps Ethernet connections to facilitate scalable AI workloads in data centers. Intel aims to offer a competitive alternative to Nvidia's dominant GPUs by leveraging an open software ecosystem in collaboration with industry partners like Google and Qualcomm.?

Gen AI for Business Trends, concerns, and predictions:?

Corporations are missing the mark by viewing AI merely as a productivity tool. Effective AI integration requires reimagining business processes and enhancing decision-making. Meanwhile, Harvard Business Review outlines four AI risks—bias, security, transparency, and ethics—and strategies to mitigate them. Generative AI is also disrupting startups, advertising, and online business models. Public opinion on AI in the news varies, with younger audiences more optimistic. Only 5% of U.S. adults use AI daily, but Indian B2B marketers are leading in AI adoption. Open-source AI poses security risks, emphasizing the need for robust safeguards. What do you think is the biggest AI challenge today??

• Corporations looking at gen AI as a productivity tool are making a mistake Corporations focusing solely on AI as a productivity tool may be making a strategic error. According to a CNBC article, the transformative potential of AI goes beyond mere productivity gains. Effective AI integration should involve reimagining business processes and enhancing decision-making. While 76% of developers plan to use generative AI, about 38% express concerns over its accuracy. AI's benefits include improved problem-solving and customer service, yet its success depends on addressing ethical, implementation, and change management challenges. By taking a holistic approach, companies can fully harness AI's potential, leading to more significant innovation and long-term value.??

• 4 Types of Gen AI Risk and How to Mitigate Them Harvard Business Review highlights four major risks associated with generative AI and offers strategies to mitigate them: bias and fairness, security and privacy, transparency and accountability, and ethical and societal impact. To address these, it suggests using diverse training data, enhancing data protection, maintaining clear documentation of AI processes, and developing ethical guidelines while engaging stakeholders. These measures aim to balance the benefits of AI with the need to manage its potential risks effectively.

• AI Chaos Is Coming for Startups, Ads and Online Business? Generative AI is poised to disrupt startups, advertising, and online businesses by transforming how web referrals and business-building activities are conducted. AI partnerships with major publishers, such as OpenAI, will change content consumption patterns, potentially creating chaos in the online economy. This shift demands that businesses adapt to new AI-driven models to maintain competitiveness, affecting everything from advertising strategies to customer engagement practices.?

• What does the public in six countries think of generative AI in news? | Reuters Institute for the Study of Journalism? A recent study by the Reuters Institute explores public opinion on generative AI in news across six countries, revealing diverse perspectives. Approximately 57% of respondents expressed concerns about the potential for AI to spread misinformation, while 35% see it as a tool to enhance news quality. Younger audiences tend to be more receptive, with 62% of those under 35 optimistic about AI's potential benefits. The study highlights the need for transparency and human oversight to maintain trust, with 74% of respondents emphasizing the importance of clear labeling of AI-generated content.?

• Did you notice that ChatGPT, Claude, and Perplexity were all down last week? Neither did I, because only about 5% of adults in the US use Gen AI on a daily basis. Exclusive: AI isn't a daily habit yet for teens, young adults? A recent survey by Axios and Generation Lab highlights the perspectives of teens and young adults on artificial intelligence (AI). The survey reveals a general optimism among young people regarding the impact of AI on their education and future careers. About 64% believe AI will enhance their learning and help them acquire necessary skills for their careers. However, concerns remain about AI's potential misuse, with 61% fearing it could facilitate cheating in school and 53% worried about privacy issues. The study also indicates a strong desire for adult guidance in navigating AI tools, with 72% of respondents expressing the need for help from adults to effectively use AI technologies.

• On another hand, in India, Three in four B2B marketing leaders in India are already using generative AI applications, says LinkedIn study - Brand Wagon News | The Financial Express? A LinkedIn study reveals that 75% of B2B marketing leaders in India are utilizing generative AI applications, with 87% planning to increase its use to boost efficiency and create engaging content. The study highlights a shift towards financial fluency and creative skills in B2B marketing, with 70% of CMOs emphasizing the importance of brand building. This trend is driven by the need to navigate economic uncertainties and prove the value of marketing investments to CFOs and business leaders. LinkedIn is also introducing new AI-powered tools like AI-generated Copy Suggestions and the CMO Scorecard to support marketers. Have you tried LinkedIn AI tools yet?

• Open source, open risks: The growing dangers of unregulated generative AI? While mainstream AI models have safety measures, open-source versions lack such protections, making them susceptible to misuse by cybercriminals for activities like phishing and generating malicious code. These vulnerabilities highlight the need for robust safeguards in AI development and emphasize that businesses adopting AI must be vigilant about internal security threats, data integrity, and the potential for model exploitation.?

What/where/how Gen AI solutions are being implemented today?

Summary

• Discord has become an unlikely center for the generative AI boom | TechCrunch? Discord, initially a chat platform for gamers, has unexpectedly become a key player in the generative AI boom. Companies like Viggle and MidJourney have leveraged Discord’s infrastructure to rapidly grow their AI communities. Viggle, which uses AI to generate controllable video content, saw its Discord membership surge from a few thousand to over 3 million within months. This growth highlights Discord's role as a scalable platform for AI startups, offering them a ready-made user base and robust content moderation tools. The collaboration benefits both parties: startups can focus on developing their AI models, while Discord handles user engagement and infrastructure challenges.

• Nvidia dominates the AI chip market, but there's more competition than ever? NVIDIA continues to dominate the AI chip market, leveraging its advanced GPUs that are critical for training and running AI models. However, competition is intensifying as companies like AMD, Intel, and emerging startups develop alternative AI hardware solutions. These competitors aim to capture a share of the rapidly growing AI hardware demand, offering innovations that challenge NVIDIA's market supremacy. This dynamic market landscape underscores the increasing importance of efficient and powerful AI computing capabilities across industries.

• Nvidia says its next-generation AI chip platform to be rolled out in 2026? NVIDIA has unveiled its next-generation AI hardware, including the Blackwell GPU and the GB200 Superchip, designed to power the emerging generative AI era. These new chips offer up to a 30x performance improvement over previous models while using significantly less energy. Key cloud providers like Amazon, Google, Microsoft, and Oracle will be among the first to integrate these technologies into their platforms. NVIDIA also introduced the DGX SuperPOD, a supercomputer system capable of supporting large-scale AI workloads with a new liquid-cooled architecture for enhanced efficiency.?

• Hi, AI: Our Thesis on AI Voice Agents | Andreessen Horowitz Andreessen Horowitz's thesis on AI voice agents highlights their potential to revolutionize phone interactions by automating calls for businesses, saving time and costs, and enhancing customer experiences. For consumers, AI voice agents can provide human-like services such as therapy and coaching. The development of these agents involves advanced models that handle speech recognition, processing, and response generation. The future of AI voice agents includes both consumer and enterprise applications, with opportunities for industry-specific solutions and integration into broader business processes.?

Regional and Regulatory Updates

The U.S. is slowing AI chip exports to the Middle East amid national security concerns as Saudi Arabia invests in China's AI sector. Executive Order 13960 mandates federal agencies to inventory AI use cases, promoting transparency and collaboration. AI is revolutionizing healthcare, with companies improving medical imaging and disease detection. U.S. regulators are probing Microsoft, OpenAI, and Nvidia for antitrust violations. The U.S. and Singapore collaborate on AI, emphasizing ethical use and workforce development. What impact will these geopolitical moves have on global AI development?

• US reportedly slows AI chip exports to Mideast as Saudi Arabia turns to China - Al-Monitor The U.S. is reportedly slowing down AI chip exports to the Middle East, including Saudi Arabia, amid a national security review. This move comes as Saudi Arabia invests in China's AI sector, notably through a $400 million investment in Chinese AI firm Zhipu AI by Saudi Aramco's venture arm. The U.S. aims to limit China's tech influence, leading to a tug-of-war over AI dominance in the region. Despite strained U.S.-China relations, Beijing is strengthening its ties with Gulf nations in AI development.?

• AI Use Cases Inventory | HHS.gov? Executive Order 13960, “Promoting the Use of Trustworthy Artificial Intelligence in the Federal Government,” requires agencies to prepare an inventory of non-classified and non-sensitive current and planned Artificial Intelligence (AI) use cases. Agencies must share their inventories with the public and other agencies, to the extent practicable. The HHS Office of the Chief Artificial Intelligence Officer (OCAIO) has created an inventory of HHS AI use cases that not only satisfies the executive order requirements, but also increases awareness of and cross-agency collaboration on AI initiatives.

• Artificial Intelligence in Health Care - BORGEN? Companies like Rology, Dawa Health, and Neural Labs Africa are leading the way by using AI to improve medical imaging, maternal and child health monitoring, and early disease detection. AI's integration in health care improves patient outcomes, reduces costs, and alleviates burdens on health systems, marking a significant advancement in modern medicine.

• US sets stage for antitrust probes into Microsoft, OpenAI and Nvidia U.S. regulators have initiated antitrust investigations into Microsoft, OpenAI, and Nvidia, focusing on their dominant roles in the artificial intelligence industry. The Justice Department will lead the probe into Nvidia, while the Federal Trade Commission (FTC) will examine Microsoft and OpenAI. These inquiries aim to assess whether these companies' practices hinder competition and innovation. Additionally, the FTC is scrutinizing Microsoft's $650 million deal with AI startup Inflection AI for potential antitrust violations. This move signals increasing regulatory scrutiny on major tech firms and their AI investments to ensure fair market practices.

• Fact Sheet: U.S.-Singapore Shared Principles and Collaboration on Artificial Intelligence | U.S. Department of Commerce? The U.S. and Singapore have announced a collaboration on AI, emphasizing shared principles like ethical AI use, transparency, and inclusivity. The partnership includes developing AI governance frameworks, boosting AI safety research, and launching the AI Talent Bridge program to enhance workforce skills. This initiative aims to leverage AI for economic growth, environmental sustainability, and societal benefits while addressing challenges through cooperative governance and standards. The collaboration also supports broader regional and international AI efforts.?

Learning Center and How To’s

• Upcoming Linkedin Lives: ethics https://www.dhirubhai.net/events/genaiforb2b-ethics7201642285619392515/comments/? and Insurance: https://www.dhirubhai.net/events/episode9-genaiforb2b-insurance7203832042113921025/comments/?

• DevOps in the era of Generative AI: Foundations of LLMOps - Microsoft Community Hub? It highlights the integration of AI in DevOps, focusing on Large Language Model Operations (LLMOps). It covers essential aspects such as continuous integration and delivery (CI/CD) for AI, monitoring and scaling AI applications, and addressing security, ethics, and governance in AI. The series aims to provide practical guidance and tools within the Azure ecosystem for effectively leveraging AI in the DevOps lifecycle.

• Reduce Your OpenAI API Costs by 70% | by Fareed Khan | Level Up Coding? To significantly reduce OpenAI API costs, you can fine-tune the GPT-3.5 model for specific tasks. Fine-tuning involves training the model further on a domain-specific dataset, allowing it to perform tasks more efficiently and at a lower cost compared to using the base GPT-4 model. This method can cut costs by over 70%, improve response speed, and maintain high-quality outputs. Effective use cases for fine-tuning include creating consistent brand voices, formatting structured outputs, and developing specialized enterprise assistants.

• How RAG helps Transformers to build customizable Large Language Models: A Comprehensive Guide - MarkTechPost? Retrieval-Augmented Generation (RAG) enhances transformer-based language models by integrating retrieval mechanisms, enabling them to access extensive external data for more accurate and relevant responses. Developed by Facebook AI, RAG combines document retrieval and generative phases to dynamically incorporate updated information without retraining the entire model. This approach improves accuracy, reduces computational resources, and offers scalability. Applications include customer support, healthcare, finance, education, and legal research, making RAG a versatile and powerful tool for developing customizable large language models.?

• From gen AI 1.5 to 2.0: Moving from RAG to agent systems | VentureBeat? Generative AI 2.0 focuses on improving long-form reasoning and understanding through enhancements like Google's Gemini 1.5, which features a million-token context window. This allows the AI to process and generate responses based on a much larger context, improving its ability to summarize extensive documents or analyze long videos comprehensively. Another major development is the Ray 2.0 framework, which simplifies the scaling and management of AI workloads, making it easier for organizations to deploy and integrate AI across various platforms. Additionally, Anthropic's Claude 2 has made strides in producing safer and longer responses, addressing concerns about harmful outputs and expanding its use cases in enterprise environments. Companies like Slack, Notion, and Zoom are leveraging Claude 2 to enhance productivity tools and customer service capabilities.

• This free chatbot lets you choose your AI model? Hugging Face has introduced HuggingChat, an open-source platform allowing users to choose large language models (LLMs) from various providers, including Meta, Google, Microsoft, and Mistral. This platform aims to provide more flexibility and customization options for AI applications. By partnering with these major tech companies, Hugging Face facilitates easier access and deployment of these models through cloud services like Microsoft Azure. This initiative underscores Hugging Face's commitment to open-source innovation, making advanced AI tools more accessible and customizable for developers.?

• Enhancing LLM Application Safety with LangChain Templates and NVIDIA NeMo Guardrails - Blockchain.News? LangChain Templates and NVIDIA NeMo Guardrails enhance the safety of large language model (LLM) applications by providing content moderation, security, and accuracy. LangChain enables swift creation of production-ready applications, while NeMo Guardrails offer programmable rules for input and output validation to prevent issues like hallucinations and cybersecurity threats. A demonstrated use case involves integrating these tools in a Retrieval-Augmented Generation (RAG) setup, ensuring secure and accurate AI responses crucial for handling sensitive data like Social Security Benefits.?

• Beginner’s Guide to Building LLM Apps with Python - KDnuggets? The beginner's guide on KDnuggets explains how to build large language model (LLM) applications using Python. It covers the essentials of setting up a Python environment, selecting appropriate pre-trained models, and using libraries like Hugging Face's Transformers for implementation. The guide provides step-by-step instructions on data preprocessing, model fine-tuning, and deploying the LLM for practical applications, including generating text and answering questions. It emphasizes the importance of understanding model limitations and ethical considerations.?

• How to Build an LLM Application With Google Gemini | HackerNoon? It involves setting up a Node.js application that acts as a chatbot, integrating the Gemini API for handling conversations, and deploying the app on Heroku. The guide includes instructions on initializing the app, obtaining an API key, and testing the application locally and on Heroku. It emphasizes the ease of use and potential applications of LLMs in various scenarios.?

Prompt of the week

Tools and Resources

• Why you need tax-specific GenAI tools despite concerns? These tools automate labor-intensive tasks such as data extraction, compliance monitoring, and tax return preparation, allowing professionals to focus on more strategic advisory roles. Among the tools mentioned are Checkpoint Edge, an AI-driven research tool offering current guidance on accounting standards and tax regulations; ONESOURCE, which manages tax processes and ensures compliance; CS Professional Suite, which streamlines accounting and tax workflows; and Onvio, a cloud-based suite facilitating comprehensive management of accounting operations.
• Resume AI Coach: https://sites.google.com/view/resumeaicoach Save loads of time using Resume AI Coach to optimize your Resume or LinkedIn profile.

If you enjoyed this newsletter, please comment and share. If you would like to discuss a partnership, or invite me to speak at your company or event, please DM me.