Stay ahead of the headlines

Mar 22, 2024·Quartz

OpenAI CEO Sam Altman, in an interview with Lex Fridman, expressed dissatisfaction with the current capabilities of GPT-4, suggesting it falls short of the company's ambitions and user expectations. Despite its shortcomings, Altman is optimistic about the future, hinting at significant improvements with the upcoming GPT-5 model, expected to be a substantial leap forward. Altman’s candid reflections underscore the exponential advancement in AI development, acknowledging GPT-4's limitation in handling complex, multi-step problems while still recognizing its utility in brainstorming scenarios. Amidst speculation about future models and the development of Sora, OpenAI's text-to-video generator, the company continues to influence the AI landscape, driven by a vision of continuous progress and innovation.

Mar 21, 2024·Reuters

Microsoft has highlighted Google's significant advantage in the generative AI sector to EU antitrust regulators, pointing out Google's vast data resources and AI-optimized chips. This competitive edge is bolstered by proprietary data from Google Search and YouTube, allowing Google to train its Gemini model more effectively. Meanwhile, Google's and Apple's dominance in AI-powered voice assistants could further strengthen their position in generative AI, disadvantaging new entrants. Microsoft emphasized the importance of pro-competitive partnerships to counteract the risk of a few companies dominating through vertical integration. Google responded by emphasizing its platform's openness compared to competitors. This discussion forms part of broader consultations by the European Commission into competition within the generative AI field.

Mar 21, 2024·UCLA Newsroom

UCLA bioengineers have developed a new AI-assisted wearable device, designed as an adhesive neck patch, to aid individuals with vocal cord dysfunctions in regaining speech capabilities. This innovative bioelectric system employs machine learning to translate laryngeal muscle movements into audible speech with nearly 95% accuracy, offering a non-invasive communication solution for those suffering from voice disorders or recovering from related surgeries. The flexible, self-powered device, detailed in the journal Nature Communications, marks a significant advancement in speech technology for disabilities, potentially transforming the communication landscape for individuals facing vocal communication challenges.

Mar 21, 2024·Fast Company

The Future Today Institute's 17th Annual Tech Trends Report, discussed by its founder Amy Webb at South by Southwest, outlines a technologically driven "supercycle" characterized by the convergence of artificial intelligence, biotech, and interconnected wearables, which promises to fundamentally reshape human civilization. This era is not just defined by technological advancements but also poses crucial socio-economic and ethical questions, such as the debate over open-source versus closed AI systems and the implications of digital twins for personal and societal change. Webb's insights underline the immediate need for strategic decisions regarding AI's deployment and highlight the potential for generative biology to revolutionize personalized medicine, suggesting that we are on the verge of transformative changes across various aspects of life.

Mar 21, 2024·CommonSpirit Health

CommonSpirit Health, a leading nonprofit health system, has introduced Insightli, its proprietary internal AI assistant, aimed at enhancing operational efficiency and data security. Launched to address the limitations of external AI solutions which often compromise privacy, Insightli enables the creation of personalized and relevant content quickly, ensuring data remains within CommonSpirit's secure environment. Senior Executive Vice President and CIO, Daniel Barchi, highlighted Insightli's success in processing over 17,000 prompts in just a few months, significantly improving workflow and allowing staff to concentrate on their core tasks. This innovative tool supports CommonSpirit's commitment to improving work lives while safeguarding sensitive information.

Mar 20, 2024·Gizmodo

Nvidia has partnered with Hippocratic AI to introduce generative AI nurses, offering healthcare providers an affordable option to address worker shortages. These AI nurses, costing $9 an hour, can provide patients with medical advice via real-time video calls, a significant saving compared to the $90 hourly rate for human nurses. Nvidia powers the technology enabling these AI agents to deliver human-like responses. The collaboration aims to alleviate the escalating nursing crisis, underscored by widespread strikes and shortages within the industry. This initiative, alongside Nvidia's other announcements at their 2024 GTC Conference, reinforces the company's influence in advancing AI technologies across various sectors.

Mar 19, 2024·AI in Healthcare

The American Medical Association (AMA), in collaboration with Manatt Health, has released a report titled "Future of Health: The Emerging Landscape of Augmented Intelligence in Health Care," which highlights eight nonclinical applications of AI (referred to as augmented intelligence) that have garnered significant interest from physicians. The applications include optimizing access to care and scheduling, streamlining the administration and revenue cycle, enhancing operational efficiency, ensuring regulatory compliance, improving patient satisfaction, promoting quality management, fostering education through feedback and training, and advancing medical research. The report emphasizes the AMA's preference for AI to augment rather than replace human decision-making in healthcare, highlighting the potential for AI tools to support physicians in various nonclinical tasks. Additionally, the need for strong data privacy measures, effective integration with current systems, and the involvement of physicians in the adoption of AI tools are key points underscored in the findings.

Mar 19, 2024·x.ai

On March 17, 2024, xAI announced the release of Grok-1, a 314 billion parameter Mixture-of-Experts language model, under the Apache 2.0 license. This groundbreaking model, distinguished by its vast parameter count and innovative architecture, was methodically trained from the ground up on an extensive corpus of text data without specific task optimization. Grok-1 stands out with 25% of its parameters being active for any given token, reflecting a strategic approach to model scaling and efficiency. The model, built using JAX and Rust, marks a significant advancement in machine learning, offering researchers and developers unparalleled access to one of the most powerful language processing tools available to date. Instructions for accessing and utilizing Grok-1's base model weights and architecture are provided on GitHub.

Mar 19, 2024·Fortune

During the Nvidia GTC Artificial Intelligence Conference in San Jose, California, Nvidia CEO Jensen Huang introduced the new Blackwell processor, designed to significantly enhance AI computing performance in both development and execution stages. Named after the pioneering mathematician David Blackwell, this chip contains 208 billion transistors and promises to revolutionize AI processing for major data center operators like Amazon, Microsoft, Google, and Oracle, with availability set for later this year. Highlighting Nvidia's leading position in AI technology and its market valuation exceeding $2 trillion, the Blackwell chip represents a leap in processing power and efficiency, incorporating innovative production techniques and offering improved networking capabilities. Huang's announcement reinforces Nvidia's central role in the ongoing AI-driven industrial transformation, emphasizing the company's strategy to democratize AI technology access across industries and government sectors.

Mar 18, 2024·Healthcare IT News

At HIMSS24, Veronica Walk from Gartner highlighted artificial intelligence (AI) and hospital-at-home technologies as key emerging trends in healthcare. AI, particularly generative AI, is swiftly advancing, offering potential to alleviate clinician burnout and enhance care team capacity by automating administrative tasks and potentially improving clinical decision-making. Meanwhile, hospital-at-home technologies, driven by workforce shortages and a desire to improve patient outcomes, are seeing significant growth, with virtual nursing emerging as a notable application. Walk emphasizes the importance for healthcare C-suite executives to not pursue technology for its own sake but to focus on solving organizational challenges and achieving specific outcomes. She advises the use of tools like Gartner's Hype Cycles to assess technologies' maturity and potential impact, and stresses collaboration among C-suite members to effectively implement and derive value from these emerging technologies.

Mar 18, 2024·Nature

Artificial Intelligence (AI) is now being used to address the slow and costly process of clinical trials, aiming to counteract the diminishing efficiency in drug development, known as Eroom's law. AI applications in trials include designing study protocols, enhancing patient recruitment, selecting eligibility criteria, and data analysis. Innovations such as predictive algorithms for trial success, automated patient recruitment systems, and digital twins to reduce control group sizes are notable examples. Furthermore, AI assists in optimizing trial parameters in real-time, identifying suitable patient candidates from databases, and employing wearables for data collection. These AI-driven methods not only expedite clinical research but also aim to broaden inclusivity in trial populations, potentially making drug development more efficient and inclusive.

Mar 18, 2024·MedCity News

At the HIMSS conference in Orlando, Hackensack Meridian Health CEO Robert Garrett emphasized AI's transformative potential in healthcare, particularly in building healthier communities at an unprecedented scale and pace. Garrett outlined four critical areas where AI could make a significant impact: increasing access to care, improving patient outcomes, advancing health equity, and addressing climate change. He highlighted examples like the partnership between Indian digital healthcare platform Apollo 24|7 and Google Cloud for AI-powered telemedicine, and Grifols' collaboration with Google Cloud to develop new plasma-based drugs. Garrett also pointed out the use of AI in early detection and treatment pathways, such as an AI tool by Carenostics for advanced kidney disease diagnosis and an app from Phoenix Children's Hospital for identifying malnutrition in pediatric patients. Additionally, Garrett spoke on the need for AI to address social determinants of health and climate change's health impacts, stressing the urgency of developing solutions that are energy-efficient and reduce pollution.

Mar 15, 2024·Beckers Hospital Review

At the HIMSS 2024 conference in Orlando, Florida, the prevailing topic was the integration of artificial intelligence (AI) in healthcare, reflecting a shift to what Stanford Health Care's Gary Fritz describes as the current "AI wave" in health technology discussions. Key highlights included the unveiling of new healthcare generative AI applications by companies like Oracle Health and Google, alongside practical deployments such as Stanford Health Care's organization-wide adoption of Microsoft’s Nuance DAX co-pilot for clinical documentation, aimed at saving physicians time. Moreover, concerns over AI safety and the importance of human oversight were discussed, as well as innovations in "Hospital at Home" programs and cybersecurity, punctuated by the recent Change Healthcare ransomware attack’s industry-wide impact. Additionally, Mayo Clinic's presentation on their future $5 billion digitized hospital campus showcased the ambitious scope of integrating digital innovation directly into patient care and operational efficiency.

Mar 15, 2024·Health Populi

The article discusses the evolving role of Artificial Intelligence (AI) in healthcare, emphasizing the importance of including patients as co-creators and end-users to ensure the technology meets their needs effectively. Dr. Grace Cordovano highlights the proactive efforts by patient communities in leveraging AI for better healthcare outcomes, challenging the industry to recognize and value their contributions. Deloitte's report on Generative AI (GenAI) in healthcare and the Pew Research findings reveal mixed sentiments among providers and consumers regarding the pace of AI adoption, underlining the potential trust and efficacy issues. The World Economic Forum (WEF) recommends a patient-first approach to GenAI, focusing on personalized healthcare, education, and management interventions to foster trust and mitigate biases. Meanwhile, initiatives like CHAI and TRAIN aim to promote responsible AI use in healthcare, addressing inclusivity and algorithmic biases. The article also highlights successful applications such as ambient listening technologies for clinical note-taking, demonstrating AI's potential to enhance patient-clinician interactions and health equity. Finally, it calls for comprehensive inclusion and responsible AI implementation, aligning with WHO's guidance on AI ethics and governance, to ensure no patient is left behind in the digital transformation of healthcare.

Mar 15, 2024·Healthcare Finance News

At the HIMSS24 conference in Orlando, healthcare leaders discussed the potential of Generative AI (GenAI) to revolutionize the sector by enhancing diagnostics, creating personalized treatment plans, and improving operational efficiencies. Despite the excitement, there was a note of caution, particularly around data security concerns and the need for responsible implementation to prevent potential data overload. Practical applications of GenAI, such as ambient listening tools to aid doctors with note-taking and AI in radiology to expedite image analysis, have already shown positive outcomes. However, panelists emphasized the importance of establishing ethical guidelines and educating leadership on both the opportunities and risks associated with AI, aiming to reduce administrative burden and enhance patient care without increasing professional burnout.

Mar 14, 2024·spectrum.ieee.org

The essence of Dina Genkina's article highlights a significant shift in the field of AI prompt engineering, particularly following the advent of ChatGPT in the fall of 2022. Initially, prompt engineering involved crafting queries to elicit the best responses from AI models, becoming a widespread practice among individuals and companies seeking to leverage AI for various applications. However, recent developments suggest that the future of prompt engineering lies in automating the process, allowing AI models to determine the most effective prompts themselves. Research by VMware on large language models (LLMs) reveals that autotuned prompts crafted by AI can outperform those devised by humans, indicating a move towards models that optimize their own prompts. This shift could render manual prompt optimization obsolete, as automated methods prove to be faster and more effective, generating unexpectedly successful and sometimes bizarre prompts. Furthermore, this approach has also been applied to image-generation algorithms, suggesting a broad potential impact across AI applications. Despite these advancements, the article suggests that the demand for human expertise in adapting AI to industry needs will persist, albeit under potentially evolving job titles and roles, reflecting the continuous and rapid evolution within the field of AI and prompt engineering.

Mar 14, 2024·San Diego Union-Tribune

In his article for the San Diego Union-Tribune, Dr. Karandeep Singh, Chief Health AI Officer at UC San Diego Health, explores the evolving role of artificial intelligence (AI) in healthcare. He began by referencing Dr. Warner Slack's work in the 1960s, which laid the foundation for computers in patient care, highlighting how AI, particularly deep learning, has followed a similar trajectory into clinical practice. Singh discusses AI's current applications at UC San Diego Health, such as the sepsis detection tool that has improved patient outcomes, and explores the potential of generative AI in personalizing care by alleviating the administrative burden on clinicians, hence fostering more patient-focused interactions. He underscores the ongoing evaluation of AI applications for safety and equity and stresses the importance of ethical AI use through UC San Diego Health's commitments and collaborations. Singh concludes by reflecting on the profound possibilities AI presents for enhancing patient safety and care quality without replacing the invaluable human element of healthcare.

Mar 14, 2024·DistilINFO HealthIT

"Revolutionizing Healthcare: Unveiling the Power of Artificial Intelligence," an article from DistilINFO Hospital IT, sheds light on the complex yet fascinating domain of AI in healthcare as presented by Dr. Jonathan Chen at the HIMSS AI in Healthcare Forum. The article encapsulates Chen's analysis of AI's potential to transform healthcare, balanced with a caution about the ethical challenges and the necessity for transparency in its application. By blending illustrative magic tricks with deep insights, Chen articulates the dual nature of AI as a tool of immense promise and peril, highlighting both its capacity to streamline healthcare operations and the ethical dilemmas it brings to the fore. The narrative stresses the importance of discernment in deploying AI, ensuring it augments rather than supplants human empathy and expertise in patient care. This piece offers a comprehensive overview of Chen's call for a nuanced approach to integrating AI in healthcare, aiming for a future where technology enhances the human aspects of care.

Mar 14, 2024·India Today

The Ministry of Electronics and Information Technology in India has issued an advisory stating that any platform training or testing an artificial intelligence (AI) tool must obtain government approval before making the product available to users within the Indian internet space. This rule is specifically aimed at large or 'significant' platforms, with an exemption for startups, as clarified by Union Minister of State for IT, Rajeev Chandrasekhar. Despite the ministry's intentions, this move has drawn criticism from tech industry leaders who see it as a potential barrier to the progress of AI technologies in India.

Mar 13, 2024·CNBC

In an article reflecting on the 35th anniversary of the World Wide Web's creation, inventor Tim Berners-Lee discusses his dissatisfaction with certain developments on the web, particularly around social media and data ownership. He forecasts a future where AI personal assistants become central to web interaction, emphasizing the need for a trustworthy digital assistant akin to doctors or lawyers. Berners-Lee envisions a shift towards users regaining control over their data through technology his startup, Inrupt, is developing, which allows for data to be stored in personal "pods." This innovation seeks to challenge the current dominance of tech giants over user data. Additionally, he touches on the prospects of mixed reality integrating seamlessly with everyday technology and speculates that regulatory pressures could lead to the breakup of a major tech company in the future, in light of recent legislation like the European Union's Digital Markets Act.