For more on programming, events and resources on Ocean AI, visit oceansupercluster.ca. This newsletter represents solely my views and not those of my employer.
- AI Unveils Whales: A team of oceanographers and marine biologists identified a mysterious noise in the Pacific Ocean as the sounds of Bryde's whales, confirmed through sightings and simultaneous recordings. Collaborating with Google, they developed an AI app to track whale movements, revealing that Bryde's whales inhabit areas near the Mariana Islands and the Pacific Ocean transition zone. This discovery may lead to new research on the unique songs of Bryde's whales. [Phys.org]
- Storm Buoy Deployment: A research program aimed at improving hurricane forecasts involves deploying scientific buoys into hurricanes' paths to collect real-time data from the ocean-atmosphere interface. During Hurricane Francine, a US Navy P-3 Orion dropped these buoys, which measure wave height, temperature, pressure, and wind speed. The data helps forecasters refine predictions on storm paths, intensity, and flooding potential. This mission highlights the importance of capturing storm details that are often missed by satellites or aircraft, enhancing understanding and prediction accuracy. [NewScientist, ShiningScience.com]
- ROV Modelling: Deep Trekker, in partnership with Qii.AI, The Department of National Defence, Kongsberg Discovery Canada Limited, and ABS Global Canada, is leading a
Canada's Ocean Supercluster
funded AI-powered ROV project to enhance ship hull inspections, improving precision, safety, and environmental protection. The project integrates cutting-edge technology like 4K cameras and multibeam sonar to create detailed 3D models and real-time data dashboards, setting new industry standards. [Marine Technology News]
- Real-Time Coastal Data: Victoria-based MarineLabs is leveraging its AI-driven marine weather data platform to enhance maritime safety and operations at the Ports of Vancouver and Prince Rupert. By providing real-time weather and coastal data, MarineLabs supports decision-making for operations like ferry scheduling and port engineering. This initiative, funded by the province, helps refine the company’s machine-learning models, improving the accuracy of coastal weather forecasts, and sets a global standard for maritime safety. [Revelstoke Review]
- Maritime Systems Partnership: Japan is exploring collaboration with AUKUS members (U.S., U.K., Australia) in maritime autonomous systems, aiming to improve interoperability. This move, part of AUKUS’s Pillar II program, leverages Japan’s expertise in AI and robotics for unmanned maritime systems, enhancing capabilities like mine countermeasures and combat support. While this partnership could strengthen Japan's defence, it also poses risks, including tensions with China and concerns about Japan's cybersecurity. The initiative reflects broader U.S. efforts to consolidate security cooperation in the Indo-Pacific region. This collaboration matters because it enhances Japan's defence capabilities and regional security, while also potentially escalating tensions with China as part of broader U.S.-led efforts to counteract Chinese influence in the Indo-Pacific. [Japan Times]
- Maritime AI: I shared this last week but it was picked up a few more times this week. At the Capital Link Shipping and Marine Services Christopher J. Wiernicki, ABS Chairman and CEO emphasized at the Capital Link Shipping and Marine Services Forum that digital twins, AI, and data are poised to revolutionize maritime safety and operations. These technologies will enhance real-time hazard detection and predictive capabilities, fundamentally redefining safety as a balance of capacity and capability against the complex demands of decarbonization and technological advancement. While the potential for safety improvements is significant, the industry must address challenges in training, systems development, and cybersecurity to fully harness these benefits. [Maritime Executive, Manila Times]
- ChatGPT's Environmental Impact: Writing a 100-word email with ChatGPT using GPT-4 consumes approximately 519 millilitres of water and 0.14 kWh of electricity, highlighting significant resource use, particularly in dry climates. If 1 million people did this weekly for a year, the AI would require over 27 million litres of water and 7,595 MWh of electricity. OpenAI acknowledges the need to improve efficiency. Note: 7,595 MWh of electricity per year could power approximately 700 average American homes annually. "It is crucial that companies deploying AI technologies remain mindful of the potential environmental costs if they fail to invest in sustainable practices, as unchecked AI growth could exacerbate global resource consumption issues" [TechRepublic]
- AI Governance Challenges: The UN’s AI advisory body released a report highlighting the fragmented and contradictory nature of global AI governance, with a focus on the risks of scaling AI without coherent regulations. The report emphasizes the need for a unified approach, proposing initiatives like international AI standards, data frameworks, and independent oversight to address these challenges. However, the influence of vested interests in AI development poses significant obstacles to meaningful regulation. Effective AI governance is crucial to preventing the amplification of harmful outcomes and ensuring responsible technology development. [TechCrunch, Forbes, The Conversation]
