AI Horizons 02 Feb 2024: Shaping Tomorrow
Exploring the Latest in AI and Machine Learning - February 2, 2024
Welcome to the February 2, 2024, edition of "AI Horizons: Shaping Tomorrow." As we delve into the vibrant world of Artificial Intelligence and Machine Learning, we aim to illuminate the path of innovation, ethics, and the profound impact these technologies have on our daily lives and the future.
Business / Finance
Amazon Introduces Rufus: AI-Powered Shopping Assistant: Amazon has launched Rufus, an AI chatbot integrated into its mobile app, designed to assist U.S. customers with product searches, comparisons, and recommendations. Trained on Amazon's vast product database and external web information, Rufus aims to enhance the shopping experience by answering queries and offering personalised advice. Initially available to a select group of users in beta, Amazon plans to expand Rufus's availability and improve its functionality based on customer feedback, ensuring a more intuitive and efficient shopping journey. Read
CitySwift Raises €7M for AI-Powered Public Transport Optimization: Galway's CitySwift secured €7 million in funding to enhance its AI platform, aiming to optimise public transportation by predicting journey times and passenger demand. This investment, led by Gresham House Ventures, will advance routing efficiency, supporting CitySwift's expansion and its mission to improve reliability and efficiency in public transport networks across major UK cities. Read
Coris Secures $3.7M for AI-Driven SMB Risk Management: California's Coris raises $3.7 million to enhance its AI risk management platform for small and medium-sized businesses (SMBs), aiming to modernise outdated financial risk evaluation processes. Backed by Lux Capital, Exponent Capital, Y Combinator, Blank Ventures, and seasoned fintech founders, Coris introduces CorShield to combat impersonation fraud, leveraging large language models for efficient data analysis and insights, meeting the growing demand for integrated, AI-embedded risk management solutions within SMB platforms. Read
Regulatory
UK's Strategic Approach to AI: Balancing Opportunities and Risks: The House of Lords report calls for a strategic rebalance in the UK government's AI policy, emphasising the need to harness the potential of large language models (LLMs) while managing immediate risks like copyright infringement and misinformation. It criticises the current focus on unlikely existential threats and advocates for a practical approach that fosters market competition to avoid regulatory capture by dominant tech firms. The report stresses the importance of openness in AI development to ensure safety, security, and accountability. It also highlights the urgency of addressing the use of copyrighted content in training LLMs and suggests that outright bans on powerful AI models are impractical. Instead, it recommends that the government's AI Safety Institute develop methods to monitor and mitigate the risks associated with AI deployment, aiming to position the UK as a leader in AI without succumbing to unfounded fears. Read
AI Ethics Code for the Insurance Industry: The insurance sector introduced an AI code of conduct developed by 127 specialists to ensure responsible AI use in claims settlements. This non-regulatory standard, led by figures like Hugh Hessing and Prathiba Krishna, emphasises fairness, accountability, and transparency, urging firms to adopt these principles to build trust in AI applications. Read
Health
AI for Suicide Prevention in Black Youth: A Collaborative Research Initiative: Researchers from the University of Texas at Austin and Cornell University, collaborating with Prairie View A&M and Tuskegee University, have received an NIH grant to develop AI interventions to reduce suicide rates among Black youth. This project, addressing a critical mental health crisis, will use natural language processing to analyse death reports and identify risk factors. The initiative emphasises ethical, responsible AI development and involves community stakeholders to ensure the relevance and trustworthiness of the interventions. Read
Tools
Google Enhances Music Creation with GenAI Tools: Google has introduced MusicFX, an advanced version of its music-generating AI, MusicLM, designed to create up to 70-second-long tracks and music loops. This tool, part of Google's AI Test Kitchen, allows users to generate music based on text prompts, offering track length and style customisation. Alongside MusicFX, Google also launched TextFX, a lyric generation tool developed with rapper Lupe Fiasco, using the PaLM 2 model to assist in the lyrics-writing process. While these tools mark Google's investment in GenAI music technology, they navigate complex issues around copyright and authenticity in AI-generated music. Read
Google Unveils ImageFX: An AI-Driven Image Creation Tool: Google introduces ImageFX, an AI-powered image generator leveraging Imagen 2, a GenAI image model from DeepMind. ImageFX allows users to create and tweak images using text prompts and "expressive chips" for enhanced creativity. Despite similarities to existing tools like DALL-E 3, ImageFX implements safety measures against misuse, including content filters and digital watermarks. Additionally, Google expands Imagen 2's application across its services, integrating it into search, ads, and AI chatbots for diverse image generation capabilities while navigating legal uncertainties around training data. Read
AI2 Releases Open-Source AI-Language Models and Training Data: The Allen Institute for AI (AI2) has launched OLMo, an open-source GenAI language model, alongside Dolma, a substantial public dataset for training. Designed for transparency in AI research, OLMo aims to enable unfettered use for training, experimentation, and commercial purposes. Unlike many "closed" models trained on proprietary data, OLMo has complete training components, offering a robust alternative for text generation tasks. AI2's commitment to openness is intended to foster ethical AI development, reproducibility, and equitable access, with plans to expand the OLMo suite with larger models and datasets. Read
Research
Advancing Carbon Nanotechnology with Machine Learning: Researchers from Tohoku University and Shanghai Jiao Tong University have developed a machine learning method to predict the growth of carbon nanostructures on metal surfaces, a breakthrough published in Nature Communications. This innovative approach simplifies the design and synthesis of carbon nanostructures, which is crucial for electronics and energy devices. The team offers a new tool for exploring carbon nanotechnology's potential without extensive experimental trials by simulating the atomic-level dynamics of graphene growth on various metals. This method paves the way for future research on advanced catalysts and energy storage materials. Read
AI and Spectroscopy Enhance Lettuce Phenotyping: A study highlights the integration of artificial intelligence and reflectance spectroscopy to improve pigment classification in lettuce, enhancing efficiency in agriculture. This approach uses AI algorithms and hyperspectral imaging for precise lettuce variety phenotyping, which is crucial for modern farming practices. The method promises advancements in crop classification and sustainable agricultural development. Read
BRAIN TEASER:
You see a boat filled with people. It has not sunk, but when you look again, you don't see a single person on the boat. Why?
MORNING CHUCKLE:
Why did the computer go to the doctor? Because it had a virus!
DID YOU KNOW:
The first AI program was written in 1951 and was designed to play a game of checkers!
As we navigate the evolving landscape of AI and ML, let's embrace the innovations and challenges they bring. Together, we're not just witnessing the future; we're actively building it. Stay curious, stay inspired, and let's shape a brighter tomorrow.