Midjourney releases Chinese version on Tencent platform.
2023-05-17
Midjourney invites Chinese creators to join beta testing on its WeChat post. The Chinese version is available on Tencent QQ, where users can apply by scanning the official QR code every Monday and Friday at 6 PM Beijing time. Users can input Chinese text to summon the AI robot for artistic creation, access the latest models, and editing features. Various themed creation channels are also available within the community.
Read more ›Apple detailed upcoming AI-driven iOS 17 accessibility features.
2023-05-17
iOS 17 to bring accessibility features for speech, vision, and cognitive impairments. Users can now type during calls and have it translated to speech. New features use machine learning. Design-oriented updates, like Assistive Access and visual media focus, aim to reduce cognitive load. Apple not alone in prioritizing accessibility, with Google and gaming industry making strides.
Read more ›iFlytek: Using Artificial Intelligence to Assist in the Fight Against "Epidemic" and Donating 10 Million Supplies
2020-02-03
iFLYTEK donated medical supplies worth 10 million yuan to support epidemic prevention efforts. It utilized AI technology to assist in epidemic control, education, and government services. The company also provided support and care for its employees during the outbreak.
Read more ›Google Open Sources LaserTagger, an AI Model for Accelerating Text Generation
2020-02-03
Google introduced the sequence-to-sequence AI model in 2014 for text generation tasks. LaserTagger, a new open-source text editing model, improves efficiency and accuracy by predicting editing operations. It is faster and more effective in various tasks, making it a significant advancement in natural language processing. LaserTagger's optimized vocabulary minimizes errors and speeds up processing, showing promising results even with limited training examples. GitHub link: https://github.com/goog
Read more ›OpenAI Fully Embraces Facebook's PyTorch Machine Learning Framework
2020-02-03
OpenAI plans to move projects to PyTorch for its efficiency and scalability. It will contribute to the community and provide educational resources. PyTorch's popularity continues to grow, with many companies migrating to it for AI projects.
Read more ›Development of a New Drug for Artificial Intelligence to Begin Clinical Trials
2020-02-03
Exscientia, an AI startup, has developed a new drug, DSP-1181, in collaboration with Sumitomo Dainippon Pharma. This AI-designed drug will undergo clinical trials in Japan in less than a year, showcasing the potential of AI-driven drug development in reducing time and costs. Despite advancements, challenges remain in ensuring the effectiveness of AI-discovered drugs and their impact on medicine.
Read more ›Application of Artificial Intelligence in Disease Transmission
2020-02-03
Artificial intelligence plays a key role in disease control. Studies on HIV, filamentous viruses, HCV, dengue fever, and TB/gonorrhea prevention highlight the effectiveness of AI in optimizing strategies and resource allocation. The fight against novel coronavirus showcases the importance of technology in overcoming challenges.
Read more ›Technology Combats Epidemic: Artificial Intelligence + Big Data
2020-02-03
The emergence of new technologies like 5G, AI, and big data is empowering traditional industries. Baidu AI offers tools for predicting novel coronavirus RNA structures, outbreak monitoring, and community outreach, aiding prevention and control efforts. Big data analysis and AI have become crucial in identifying infection sources and managing cases efficiently during the ongoing epidemic. Efforts like Baidu's intelligent outbound platform and Alibaba's "Intelligent Epidemic Robot" are leveraging
Read more ›IBM Launches Watson Studio, Making Deployment of Generative AI Easier
2023-05-12
IBM announced Watsonx, a platform for training, fine-tuning, and deploying AI models. It includes three product sets: watsonx.ai, watsonx.data, and watsonx.governance. Customers can build custom AI models using tools like fm.code and fm.geospatial. IBM is also collaborating with Hugging Face to integrate open-source AI models. IBM plans to incorporate Watsonx AI into all major software products to enhance IT operations, workforce solutions, and environmental risk management.
Read more ›Artificial Intelligence Accelerating Discovery of Celestial Bodies
2023-05-15
AI is revolutionizing astronomy, improving image clarity, classifying galaxies with 98% accuracy, searching for exoplanets, and even predicting new astronomical phenomena. As telescope data grows exponentially, AI is essential for handling the vast datasets and making new cosmic discoveries.
Read more ›Google Bard AI Officially Launched
2023-05-15
Google Bard, a direct competitor to ChatGPT, is now fully available. It runs on Google's language model and integrates with Google Search. Users can access Bard at bard.google.com. Google aims to expand Bard to other languages and add features like code writing. User feedback is crucial for improvement. Google competes with Microsoft and OpenAI in the AI assistant market. Leaders should consider AI's impact on productivity and job availability.
Read more ›The Impact of Artificial Intelligence on the Fashion Industry
2023-05-15
Overview: AI impacts fashion, threatens traditional weavers & artisans. AI's job automation trend raises economic concerns. Fashion industry faces sustainability challenges with AI adoption. Sustainable fashion requires investment & consumer education. Luxury brands experimenting with AI raise prices, exacerbating economic inequality.
Read more ›Artificial Intelligence Can Play Games Like Humans
2020-01-31
The Uber AI Research team studied if artificial intelligence can play text-based games like humans. They developed a system based on exploration and imitation learning to solve these games. The model maps observed outcomes to actions, exploring the environment and recording visited locations. In experiments, the model found an optimal strategy, reducing actions by half compared to previous systems, while still maintaining a trajectory. This method shows promise in large action games.
Read more ›Enriching Digital Maps with Artificial Intelligence
2020-02-01
Researchers from MIT and Qatar Computing Research Institute use satellite images for road labeling in digital maps to enhance GPS navigation. RoadTagger, a model presented at AAAI, predicts lane counts and road types with 77% and 93% accuracy, respectively. It utilizes a combination of neural network architectures, including graph neural networks, to extract road features from obscured objects like trees and buildings.
Read more ›Revealing the Accuracy of Patient Risk Models
2020-02-02
Risk models used after heart attacks or strokes can be inaccurate. MIT researchers developed a method to determine model trustworthiness, helping doctors choose better treatments and avoid unnecessary risks.
Read more ›Learning from Mistakes: Worker Robots
2020-01-31
Computer scientists at the University of Leeds are using AI for automated planning and reinforcement learning to train robots in complex environments. Their goal is to develop autonomous robots capable of problem-solving and manipulation. This research combines planning and reinforcement learning to improve robot decision-making speed by 10 times.
Read more ›Call for Suspension of Facial Recognition Highlights the Necessity of Preventing Surveillance Technology
2020-02-01
The debate over facial recognition regulation continues, with Google's CEO supporting a temporary suspension. Microsoft, however, emphasizes the technology's positive impact. While some call for delays, AI industry investments remain strong, with concerns over surveillance and privacy.
Read more ›Caution Needed in the Detection of Cancer by Artificial Intelligence
2020-02-02
Algorithms play a key role in healthcare, but digitizing nuanced judgment skills is a challenge. AI's cancer detection capabilities can lead to overdiagnosis, raising concerns about the need for uncertainty in medical diagnostics. The fusion of AI and medicine has sparked debate over the role of radiologists and the limitations of deep learning in complex medical cases. Amidst the hype, experts stress the importance of cautious deployment of AI in healthcare to avoid unintended consequences.
Read more ›Google Introduces Cross-Platform AI Pipeline Framework MediaPipe to the Web
2020-01-31
Google released MediaPipe for AI pipelines with object detection, face tracking, etc. It now supports WebAssembly, XNNPack, and WebGL for web deployment, allowing real-time webcam testing in the browser at viz.mediapipe.dev. It has limitations like no TensorFlow Lite inference and single-thread graph computation. The team is working to address these issues for more developer control.
Read more ›Universal Super Cruise now has the ability to automatically change lanes.
2020-02-01
General Motors updates Super Cruise for Cadillac CT5, CT4, and Escalade, offering improved functionality and lane-changing capabilities on highways. Improvements include better user interface, rearward sensors, and enhanced driver attention monitoring system. System now has over 130,000 miles of compatibility in the US and Canada, with plans to expand to other GM brands after 2020.
Read more ›