• AI KATANA
  • Posts
  • New AI tool can diagnose cancer, guide treatment, predict patient survival

New AI tool can diagnose cancer, guide treatment, predict patient survival

Also: GPT Next Marks AI Milestone, OpenAI Japan’s Tadashi Nagasaki Reveals

Author

AI KATANA
September 04, 2024

Good morning,

Today’s edition brings exciting developments from the cutting edge of AI, spanning breakthroughs in healthcare, major funding milestones, and innovative artistic ventures. We begin with a revolutionary AI tool developed by Harvard Medical School that is set to transform cancer diagnostics and personalized treatment. Next, OpenAI Japan has revealed its latest milestone as it introduces “GPT Next,” which promises to dramatically enhance the capabilities of AI across industries. We also cover the $100M funding round for Sakana AI, positioning it as a key player in Japan’s AI landscape, with backing from Nvidia. On the creative front, Refik Anadol’s Echoes of the Earth exhibition in Seoul offers a stunning fusion of AI and nature, using technology to highlight the fragility of ecosystems. Finally, OpenAI co-founder Ilya Sutskever’s new venture, Safe Superintelligence (SSI), has raised $1 billion, with a focus on ensuring AI develops safely and ethically. As AI continues to evolve rapidly, these stories provide a glimpse into the future of technology across healthcare, art, and industry.

Sliced just for you:

  • 🧬 New AI tool can diagnose cancer, guide treatment, predict patient survival

  • 🤖 GPT Next Marks AI Milestone, OpenAI Japan’s Tadashi Nagasaki Reveals

  • 💸 Nvidia Backs Sakana AI in $100M Funding Round

  • 🎨 Refik Anadol’s Echoes of the Earth: Living Archive at FUTURA Seoul – A Journey Through the Convergence of AI and Nature

  • 💼 OpenAI Co-Founder Ilya Sutskever’s New AI Firm Raises $1 Billion

Researchers at Harvard Medical School have developed a groundbreaking AI tool named CHIEF (Clinical Histopathology Imaging Evaluation Foundation) that can diagnose cancer, guide treatment decisions, and predict patient survival across 19 different types of cancer. This versatile AI model was trained on millions of images and tested on thousands of tissue samples from 24 hospitals worldwide. Unlike many current AI systems, which are limited to specific cancer types or tasks, CHIEF excels in detecting cancer cells, predicting tumor molecular profiles, and forecasting treatment responses with remarkable accuracy. It also outperformed existing AI models in identifying genetic mutations linked to treatment efficacy and predicting patient outcomes, offering insights into the tumor microenvironment that were previously unknown. CHIEF’s ability to work with both biopsy and surgical samples across different clinical settings represents a major advancement in AI-driven cancer diagnostics, with the potential to revolutionize personalized cancer treatment and global healthcare accessibility.

At the KDDI Summit 2024, OpenAI Japan’s representative Tadashi Nagasaki announced that ChatGPT had surpassed 200 million global users by the end of August 2024, showcasing the deep integration of AI into daily life and business operations. Alongside this milestone, Nagasaki introduced “GPT Next,” the upcoming iteration in OpenAI’s series of models, which is expected to be up to 100 times more advanced than GPT-4. While details remain scarce, GPT Next is anticipated to greatly enhance multimodal capabilities, enabling even more complex interactions with various data types. The model is poised to revolutionize industries such as healthcare, finance, and creative sectors by offering more intuitive and accurate AI tools. This marks a new era in AI development, with GPT Next set to drive innovation and reshape the way businesses and individuals engage with technology. As AI evolves at an exponential rate, its potential applications are only beginning to be realized.

Sakana AI, a Japanese AI startup, has raised over $100 million in a Series A funding round, with Nvidia participating both as an investor and a strategic partner. This funding round, led by New Enterprise Associates, Khosla Ventures, and Lux Capital, positions Sakana AI to challenge global AI leaders like OpenAI and Anthropic. The company’s unique approach, focused on nature-inspired intelligence and evolutionary optimization, aims to push AI innovation forward. Through its partnership with Nvidia, Sakana AI will have access to advanced GPU technologies and Japan-based data centers to further develop its models and infrastructure. Nvidia is also supporting Sakana AI’s efforts to build Japan’s AI community through initiatives like hackathons and outreach programs. Sakana AI’s vision includes addressing Japan’s societal challenges and fostering the development of AI models that reflect the country’s culture and values, making it a strong contender in the global AI race.

Refik Anadol’s exhibition Echoes of the Earth: Living Archive at FUTURA Seoul presents a captivating fusion of AI and nature, offering a deeply immersive experience that reimagines the natural world. At the heart of the exhibition is the Large Nature Model (LNM), an open-source AI model developed by Anadol’s studio, trained on a decade’s worth of nature data from institutions like the Smithsonian Museum. The model generates stunning visual art that captures the complexity and beauty of ecosystems, such as tropical rainforests, using data-driven storytelling to highlight environmental fragility. The exhibition also includes multi-sensory elements like AI-generated scents, enhancing the immersive experience. Standout pieces like Machine Hallucinations – LNM Fauna and Artificial Realities: Coral explore biodiversity through AI-generated interpretations of wildlife and coral, raising awareness of ecological crises. Anadol’s work exemplifies how AI can deepen our connection to nature while posing essential questions about technology’s role in environmental preservation. Running from September 5th to December 8th, 2024, the exhibition offers a visionary glimpse into the convergence of art, AI, and the natural world.

💼 OpenAI Co-Founder Ilya Sutskever’s New AI Firm Raises $1 Billion

Ilya Sutskever, co-founder of OpenAI, has launched a new venture, Safe Superintelligence (SSI), which has raised $1 billion in funding just three months after its inception. Backed by major investors such as Andreessen Horowitz, Sequoia Capital, and DST Global, SSI focuses on developing “safe superintelligence”—AI systems that exceed human intelligence while remaining aligned with human values. With offices in Palo Alto and Tel Aviv, the company is assembling a specialized team to prioritize AI safety and responsible development, distinguishing itself from competitors like OpenAI and Anthropic. The funding will support SSI’s research and development efforts as it aims to prevent the risks associated with advanced AI systems. Sutskever’s departure from OpenAI was driven by disagreements over its direction, particularly the emphasis on product development over long-term AI safety, leading him to pursue this new, safety-centric approach.

🧑🏽‍💻 AI Jobs

🛠️ AI tools updates

A new AI tool called xFakeSci, developed by a researcher at Binghamton University, has demonstrated a 94% accuracy in detecting fake scientific articles. This tool addresses a growing issue in the scientific community, where AI-generated papers are increasingly difficult to distinguish from legitimate research. xFakeSci analyzes specific patterns in writing, such as the frequency of word pairs (bigrams) and how they are connected, to identify discrepancies between genuine and AI-generated content. Initial tests focused on medical topics like Alzheimer’s, cancer, and depression, revealing significant differences in the structure of fake articles. The tool outperforms existing methods, nearly doubling the success rate of detecting fraudulent research. As the prevalence of AI-generated content rises, xFakeSci is a crucial step toward preserving the integrity of scientific literature and preventing the spread of misinformation in critical fields like medicine. The team aims to expand the tool’s capabilities across other disciplines to ensure its effectiveness in a broader range of research areas.

💵 Venture Capital updates

South Korean startups are forging stronger partnerships with global companies such as LVMH, OpenAI, and LG to scale their innovative technologies. MarqVision, a startup specializing in AI-powered intellectual property protection, has expanded globally with support from LVMH, participating in its accelerator program. Healthcare startup Exosystems is collaborating with Roche to develop devices for musculoskeletal disorders, while sleep tracking technology company Asleep is working with OpenAI to integrate its service into ChatGPT. These collaborations are part of a broader effort by both startups and corporations to leverage disruptive technologies for growth. The Korean government is also supporting these initiatives, inviting international tech giants like Nvidia and Microsoft to work with local startups through programs that have led to significant investments and partnerships. Despite some challenges in corporate venture capital funding, these partnerships represent a growing trend of collaboration between South Korean startups and multinational companies.

🫡 Meme of the day

⭐️ Generative AI image of the day