AI Innovations
1. GenStudio
Adobe has unveiled its GenStudio platform, an AI-driven tool designed to streamline the creation of marketing campaigns, revealed at the company’s Summit event. GenStudio provides an integrated environment with generative AI features for producing and optimizing ads across various platforms, aiming to deliver efficient campaign management and asset generation. Alongside GenStudio, Adobe is enhancing its AI offerings with updates to Firefly, a new AI assistant for the Experience platform, and a collaboration with Microsoft to integrate Experience Cloud with Microsoft 365 (The Verge).
2. Sanaka AI
Sakana AI, established by ex-Google researchers David Ha and Llion Jones, has developed a novel approach to AI model development, leveraging evolutionary techniques and collective intelligence to merge various open-source models. Their unique method, Evolutionary Model Merge, automates the integration of diverse models, yielding advanced capabilities and spawning models with new functionalities, as demonstrated by their success with a Japanese Large Language Model and a Vision-Language Model (MSN).
3. Ray-Ban
Meta is set to enhance its Ray-Ban smart glasses with AI capabilities next month, introducing features such as object, animal, and monument recognition, as well as language translation. The glasses will operate via a voice command system, enabling users to interact with a smart assistant integrated within the frames for various functions (The Verge).
4. DBRX
Databricks recently unveiled DBRX, an advanced open-source large language model, outperforming previous models like Meta’s Llama 2 and showcasing superior capabilities in various AI benchmarks (Wired).
5. Hume AI
Hume AI, a startup based in New York, has secured $50 million in funding to develop an advanced AI capable of processing and responding to human emotions through a voice-based interface, aiming to make interactions more intuitive and applicable in various settings like healthcare. Their technology, called the empathetic voice interface (EVI), is designed to understand emotional cues from spoken words to tailor responses, marking a significant step beyond traditional text-based AI systems (Inc).
6. AI for Travel Itineraries
Google is enhancing its travel planning process with a new AI-powered feature within its Search Generative Experience (SGE), allowing users to create personalized travel itineraries by simply making requests in search, such as planning a multi-day trip focusing on specific interests. This innovative feature, available in English and currently limited to the U.S., draws on a vast array of web data, including reviews and user-submitted information, and integrates seamlessly with other Google services like Gmail, Docs, and Maps (CNET).
7. NVIDIA
Nvidia is making significant strides in healthcare with the introduction of around two dozen AI-powered tools at its 2024 GTC AI conference, including partnerships with Johnson & Johnson for AI in surgery and GE Healthcare for medical imaging enhancements. The company’s decade-long development in healthcare AI aims to transform areas like drug discovery, which is traditionally costly and time-consuming, by leveraging AI to streamline processes and enhance precision in medical applications (CNBC).
8. Improving complex beer flavor
By analyzing chemical properties, sensory data, and consumer reviews of 250 beer varieties, researchers trained machine learning models to predict beer flavors and consumer preferences, with Gradient Boosting models outperforming traditional statistical approaches. The study highlights the potential of machine learning to identify key flavor compounds and guide the creation of beers with enhanced consumer appeal, demonstrating a novel approach to understanding and improving food flavor complexity (Nature Communications).
9. Grok-1.5
Elon Musk announced the upcoming release of xAI’s new chatbot, Grok-1.5, on his social media platform X, with plans for a more advanced Grok 2 currently in development. Following a lawsuit against OpenAI, Musk’s xAI aims to rival major AI players by enhancing its chatbot offerings and committing to open-source principles (Reuters).
10. AI fact check
SAFE, an AI system, employs a large language model (LLM) to deconstruct responses into individual facts and assess their accuracy by comparing with Google Search results, demonstrating a 72% alignment with human raters and outperforming them in a majority of disputes. Critics argue that labeling SAFE’s performance as “superhuman” may be misleading without comparing it to expert fact-checkers, emphasizing the need for a clear benchmark (VentureBeat).
11. Voice generation
OpenAI has introduced Voice Engine, a text-to-voice AI that can generate a synthetic voice from just a 15-second audio sample, currently utilized in limited applications including ChatGPT’s Read Aloud feature. This technology, while in its nascent stage, is being tested with select companies and includes strict usage policies to prevent misuse, such as impersonation or unauthorized voice cloning, while promoting transparency and consent in its deployment (The Verge).
12, AI flight attendant
Qatar Airways has introduced Sama 2.0, a virtual AI flight attendant designed to provide real-time customer assistance both at airports and on flights, offering a blend of technology and personalized service. This innovative digital crew member, accessible through various digital platforms, aims to enhance passenger experiences with interactive guidance and support, marking a significant advancement in integrating AI into customer service in the aviation industry (New York Post).
Other Innovations
1. Epilepsy treatment
Neurona Therapeutics is advancing stem-cell technology with an experimental treatment for epilepsy, where patients, including Justin Graves, receive transplants of lab-made neurons into their brains, showing promising early results in reducing seizures. This innovative approach, which injects “inhibitory interneurons” to calm brain activity, represents a potential shift from traditional epilepsy surgeries that remove or destroy brain tissue, offering a less invasive and potentially restorative alternative (MIT Tech Review).
2. Wearable
Researchers have developed a smart adhesive patch that effectively channels sweat away from its central electronics, ensuring continuous and accurate monitoring of biometrics or environmental conditions without damage from moisture. This innovation in wearable technology addresses the challenge of integrating sweat-wicking properties into small, flexible patches containing sensitive sensors and circuits, which are prone to malfunction due to moisture-induced short circuits and corrosion. The device is flexible, adheres securely to the skin, and efficiently expels sweat, enhancing wearer comfort and ensuring the reliability of the data collected over extended periods (Nature).
3. XB-1
Boom Supersonic has successfully conducted the inaugural flight of XB-1, the world’s first independently developed supersonic jet, marking a significant milestone toward the revival of supersonic air travel. The XB-1 demonstrator aircraft incorporates advanced technologies like carbon fiber composites, digitally optimized aerodynamics, and a supersonic propulsion system, validating key innovations for Boom’s commercial supersonic airliner, Overture. With major airlines placing orders for Overture, Boom Supersonic is paving the way for a new era of faster, more sustainable air travel (PR Newswire).
Articles
1. AI chatbots here to help with your mental health, despite limited evidence they work (ABC News)
AI chatbots are increasingly marketed as solutions to the mental health crisis among teens and young adults, offering support through interactive conversations and therapeutic exercises without claiming to provide formal therapy. While these chatbots offer accessible and stigma-free support, there is scant evidence of their effectiveness in improving mental health, and none have undergone FDA approval processes to validate their efficacy in treating conditions like depression. The debate continues on whether these tools should be more rigorously regulated, as they begin to play a more significant role in mental health care due to the shortage of professionals, despite concerns about their long-term impact and ability to handle severe mental health issues.
2. What does AI mean for a responsible business?(Financial Times)
While AI can drive innovations across sectors and enhance efficiency, irresponsible use of AI could lead to societal harm, data breaches, and job losses. Companies must implement robust governance frameworks, prioritize ethical AI practices, and collaborate with stakeholders to harness AI’s potential responsibly. Investors are also scrutinizing how companies use AI, especially from an environmental, social, and governance (ESG) perspective. Policymakers worldwide are taking steps to regulate AI, recognizing the need to mitigate its risks while enabling innovation.
3. Inkjets are for more than just printing (IEEE Spectrum)
Inkjet technology, originally developed for printing on paper, has found diverse applications beyond its original purpose, enabling precise deposition of microscopic droplets of various materials. This versatile technology is now being utilized in fields such as genomics for creating DNA microarrays, 3D printing for manufacturing intricate structures, and personalized medicine for producing customized drug formulations. With its ability to pattern a wide range of materials with high precision and flexibility, inkjet technology continues to open up new possibilities across multiple domains.
4. Memories are made by breaking DNA — and fixing it (Nature)
A new study on mice reveals that the formation of long-term memories involves the breaking and subsequent repairing of DNA strands in certain brain cells, accompanied by an inflammatory response. This unexpected process, where neurons seemingly treat their own damaged DNA as a signal from an invader, suggests that memory formation might be encoded through cycles of DNA damage and repair. The findings provide insights into the cellular mechanisms underlying memory formation and persistence, potentially shedding light on neurodegenerative diseases where such processes could be impaired.
