AI News(05/19): Google I/O unveils an agent that never sleeps, while scientists build living AI brains
Today, the artificial intelligence landscape is a whirlwind of innovation, showcasing remarkable advancements in both digital and biological realms. Google's I/O conference took center stage, unveiling powerful new models and a groundbreaking always-on agent, while breakthroughs in synthetic biology and massive infrastructure investments highlight AI's accelerating trajectory.
Google's I/O announcements: new models, a cloud agent that never sleeps, and a redesigned Gemini app
Google's I/O developer conference delivered a flurry of AI news, led by the introduction of Gemini 3.5 Flash, a model touted for its speed, cost-effectiveness, and agentic capabilities, outperforming its predecessors on key benchmarks. The company also unveiled Gemini Omni, a truly multimodal model designed to process any input to produce any output, starting with video. Perhaps most intriguing is Gemini Spark, a personal agent running 24/7 in the cloud, integrated with Workspace and promising constant assistance even when devices are off. The Gemini app also received a significant refresh. Why it matters: Google is making a bold play for agentic AI leadership, emphasizing speed, cost efficiency, and persistent, integrated AI assistance for both developers and end-users.
Scientists Are Building Artificial Brains from Living Cells
Princeton University researchers are pushing the boundaries of AI hardware by creating 3D-MIND, a hybrid system that integrates tens of thousands of living rat hippocampal neurons into a flexible 3D polymer mesh with electrodes. This biological neural network (BNN) aims to mimic the complexity and energy efficiency of the human brain, addressing the growing power demands of conventional AI. Why it matters: This research offers a radical alternative to silicon-based AI, potentially unlocking unprecedented energy efficiency and complexity for future computing.
Blackstone to invest $5 billion in AI infrastructure venture with Google, powered by TPU chips
Blackstone, the world's largest private owner of data centers, announced a $5 billion equity investment in a new AI infrastructure company in partnership with Google. This venture will utilize Google's purpose-built Tensor Processing Units (TPUs), bringing 500 megawatts of compute capacity online by 2027. This move intensifies Google's competition with Nvidia in the AI hardware market. Why it matters: A massive investment underscores the surging demand for specialized AI compute infrastructure and Google's strategic push to expand its proprietary TPU ecosystem.
Indian scientists create world’s first AI-designed gene editor for crops
Scientists at the Indian Council of Agricultural Research's Central Rice Research Institute have developed Plant-OpenCRISPR1 (POC1), the world's first AI-designed genome-editing tool for plants. This breakthrough moves beyond reliance on naturally occurring microbial proteins (like traditional CRISPR systems) and offers a versatile, potentially cheaper, and patent-unburdened method for enhancing crop traits like yield and climate resilience. Why it matters: This innovation could revolutionize agricultural biotechnology, offering novel solutions for food security and crop improvement with broader accessibility.
InterSystems IntelliCare Becomes the First AI-Native EHR to Achieve EU Medical Device Regulation Certification
InterSystems announced that its IntelliCare and TrakCare electronic health record (EHR) solutions have achieved EU Medical Device Regulation (MDR) Class IIa certification. This marks the first fully unified AI-native EHR to secure such approval in the EU, signifying that its AI capabilities are fundamentally designed into the platform’s data layer, not merely bolted on, ensuring stringent safety and quality standards. Why it matters: Sets a critical precedent for responsible AI integration in healthcare, ensuring compliance and building trust among providers and regulators for AI-powered clinical tools.
How AI helped treat a newborn’s ultra rare disease. ‘It was almost like a light switch.’
The parents of Jorie Kraus shared a powerful testimony at STAT's Breakthrough Summit West, detailing how AI played a crucial role in diagnosing and treating their daughter’s ultra-rare disease. The intervention was described as a pivotal moment, akin to a "light switch" for their child's condition. Why it matters: Highlights AI's profound and often life-saving impact in pediatric medicine, especially for diagnosing and treating highly complex and rare genetic conditions.
Gemini 3.5: frontier intelligence with action
Google's official blog post elaborates on Gemini 3.5 Flash, emphasizing its "frontier intelligence with action." It highlights the model's exceptional speed—four times faster than other frontier models—and its ability to tackle complex, long-horizon agentic tasks with high performance at a fraction of the cost. The integration with Google's Antigravity harness further empowers it for demanding, collaborative subagent workflows. Why it matters: Reinforces Google's strategic focus on developing highly efficient, agentic AI capable of autonomous multi-step problem-solving, poised to transform various industries.
Apples AI-powered accessibility upgrades include eye-controlled wheelchairs
Apple unveiled a suite of new AI-powered accessibility features for its ecosystem, coinciding with Global Accessibility Awareness Day. These upgrades include enhanced VoiceOver descriptions, natural language Voice Control for device navigation, on-device automatic video captioning, and a groundbreaking feature allowing power wheelchair users to navigate their devices using only an Apple Vision Pro headset with eye-tracking. Why it matters: Apple's commitment to integrating AI into accessibility tools significantly empowers individuals with disabilities, making technology more intuitive and inclusive.
Simulate real-world places with Project Genie and Street View
Google is expanding access to Project Genie, its general-purpose world model, to Google AI Ultra subscribers globally. A new capability allows users to ground generative AI environments in real-world Street View imagery. This enables creating interactive virtual worlds based on actual locations, allowing AI agents or users to explore or reimagine places with various stylistic twists. Why it matters: This advancement bridges generative AI with real-world geographical data, creating richer, more practical simulation environments for AI development and immersive user experiences.
Standard Chartered to replace ‘lower-value human capital’ with AI
Standard Chartered has announced plans to replace "lower-value human capital" with artificial intelligence. While specific details were not fully provided in the report, this statement signals a strategic shift in the bank's operational model, leveraging AI to optimize its workforce and processes. Why it matters: Illustrates the accelerating trend of AI-driven workforce transformation, posing significant questions about job displacement and the future of work in various sectors.
To stay up to date on everything going on in AI, check out the tracker at the500feed.com
Sources
- Google's I/O announcements: new models, a cloud agent that never sleeps, and a redesigned Gemini app
- Scientists Are Building Artificial Brains from Living Cells
- Blackstone to invest $5 billion in AI infrastructure venture with Google, powered by TPU chips
- Indian scientists create world’s first AI-designed gene editor for crops
- InterSystems IntelliCare Becomes the First AI-Native EHR to Achieve EU Medical Device Regulation Certification
- How AI helped treat a newborn’s ultra rare disease. ‘It was almost like a light switch.’
- Gemini 3.5: frontier intelligence with action
- Apples AI-powered accessibility upgrades include eye-controlled wheelchairs
- Simulate real-world places with Project Genie and Street View
- Standard Chartered to replace ‘lower-value human capital’ with AI