Harvard Welcomes NTT Research’s New Physics of Artificial Intelligence Group

Decoding the Mystery of Artificial Intelligence: A Closer Look at the Black Box Problem

Understanding AI Through Associations and Patterns

When a parent is teaching their young child to relate to the world, they teach through associations and the identification of patterns. Take the letter S, for example. Parents show their child enough examples of the letter and before long, they will be able to identify other examples in contexts where guidance is not active; school, a book, a billboard.

The Emergence of the Black Box Problem in AI

Much of the ever-emerging artificial intelligence (AI) technology was taught the same way. Researchers fed the system correct examples of something they wanted it to recognize, and like a young child, AI began recognizing patterns and extrapolating such knowledge to contexts it had never before experienced, forming its own “neural network” for categorization. Like human intelligence, however, experts lost track of the inputs that informed AI’s decision making. 

Establishing Trust and Safety in AI Systems

The “black box problem” of AI thus emerges as the fact that we don’t fully understand how or why an AI system makes connections, nor the variables that play into its decisions. This issue is especially relevant when seeking to improve systems’ trustworthiness and safety and establishing the governance of AI adoption. 

The Launch of the Physics of Artificial Intelligence Group

Now, a new independent study group will address these challenges by merging the fields of physics, psychology, philosophy and neuroscience in an interdisciplinary exploration of AI’s mysteries.

  1. What is the Physics of Artificial Intelligence Group at Harvard?
    The Physics of Artificial Intelligence Group at Harvard is a new research group launched by NTT Research, focusing on the intersection of physics and AI.

  2. What is the goal of the Physics of Artificial Intelligence Group at Harvard?
    The goal of the group is to explore and apply principles from physics to improve the understanding and development of AI technologies.

  3. How will the group’s research benefit the field of artificial intelligence?
    By incorporating insights from physics, the group aims to enhance the efficiency, robustness, and capabilities of AI systems, leading to advancements in various applications and industries.

  4. Who will be leading the research efforts of the Physics of Artificial Intelligence Group at Harvard?
    The group will be led by Professor Hopfield, a renowned physicist and AI expert, along with a team of researchers and collaborators from Harvard and NTT Research.

  5. How can individuals or organizations get involved with the Physics of Artificial Intelligence Group at Harvard?
    Interested parties can reach out to NTT Research or Harvard University to learn more about potential collaborations, partnerships, or opportunities to support the group’s research initiatives.

Source link

NTT Introduces Revolutionary AI Inference Chip for Instantaneous 4K Video Processing on the Edge

NTT Corporation Unveils Groundbreaking AI Inference Chip for Real-Time Video Processing

In a significant advancement for edge AI processing, NTT Corporation has introduced a revolutionary AI inference chip capable of processing real-time 4K video at 30 frames per second while consuming less than 20 watts of power. This cutting-edge large-scale integration (LSI) chip is the first of its kind globally to achieve high-performance AI video inferencing in power-constrained environments, marking a breakthrough for edge computing applications.

Bringing AI Power to the Edge: NTT’s Next-Gen Chip Unveiled

Debuted at NTT’s Upgrade 2025 summit in San Francisco, this chip is designed specifically for deployment in edge devices, such as drones, smart cameras, and sensors. Unlike traditional AI systems that rely on cloud computing for inferencing, this chip delivers potent AI capabilities directly to the edge, significantly reducing latency and eliminating the need to transmit ultra-high-definition video to centralized cloud servers for analysis.

The Significance of Edge Computing: Redefining Data Processing

In the realm of edge computing, data is processed locally on or near the device itself. This approach slashes latency, conserves bandwidth, and enables real-time insights even in settings with limited or intermittent internet connectivity. Moreover, it fortifies privacy and data security by minimizing the transmission of sensitive data over public networks, a paradigm shift from traditional cloud computing methods.

NTT’s revolutionary AI chip fully embraces this edge-centric ethos by facilitating real-time 4K video analysis directly within the device, independent of cloud infrastructure.

Unlocking New Frontiers: Real-Time AI Applications Redefined

Equipped with this advanced chip, a drone can now detect people or objects from distances up to 150 meters, surpassing traditional detection ranges limited by resolution or processing speed. This breakthrough opens doors to various applications, including infrastructure inspections, disaster response, agricultural monitoring, and enhanced security and surveillance capabilities.

All these feats are achieved with a chip that consumes less than 20 watts, defying the hundreds of watts typically required by GPU-powered AI servers, rendering them unsuitable for mobile or battery-operated systems.

Breaking Down the Chip’s Inner Workings: NTT’s AI Inference Engine

Central to the LSI’s performance is NTT’s uniquely crafted AI inference engine, ensuring rapid, precise results while optimizing power consumption. Notable innovations include interframe correlation, dynamic bit-precision control, and native YOLOv3 execution, bolstering the chip’s ability to offer robust AI performance in once-constrained settings.

Commercialization and Beyond: NTT’s Vision for Integration

NTT plans to commercialize this game-changing chip by the fiscal year 2025 through NTT Innovative Devices Corporation. Researchers are actively exploring its integration into the Innovative Optical and Wireless Network (IOWN), NTT’s forward-looking infrastructure vision aimed at revolutionizing modern societal backbones. Coupled with All-Photonics Network technology for ultra-low latency communication, the chip’s local processing power amplifies its impact on edge devices.

Additionally, NTT is collaborating with NTT DATA, Inc. to merge the chip’s capabilities with Attribute-Based Encryption (ABE) technology, fostering secure, fine-grained access control over sensitive data. Together, these technologies will support AI applications necessitating speed and security, such as in healthcare, smart cities, and autonomous systems.

Empowering a Smarter Tomorrow: NTT’s Legacy of Innovation

This AI inference chip epitomizes NTT’s commitment to fostering a sustainable, intelligent society through deep technological innovation. As a global leader with a vast reach, NTT’s new chip heralds the dawn of a new era in AI at the edge—a realm where intelligence seamlessly melds with immediacy, paving the way for transformative advancements in various sectors.

  1. What is NTT’s breakthrough AI inference chip?
    NTT has unveiled a breakthrough AI inference chip designed for real-time 4K video processing at the edge. This chip is able to quickly and efficiently analyze and interpret data from high-resolution video streams.

  2. What makes this AI inference chip different from others on the market?
    NTT’s AI inference chip stands out from others on the market due to its ability to process high-resolution video data in real-time at the edge. This means that it can analyze information quickly and provide valuable insights without needing to send data to a centralized server.

  3. How can this AI inference chip be used in practical applications?
    This AI inference chip has a wide range of practical applications, including security monitoring, industrial automation, and smart city infrastructure. It can help analyze video data in real-time to improve safety, efficiency, and decision-making in various industries.

  4. What are the benefits of using NTT’s AI inference chip for real-time 4K video processing?
    Using NTT’s AI inference chip for real-time 4K video processing offers several benefits, including faster data analysis, reduced latency, improved security monitoring, and enhanced efficiency in handling large amounts of video data.

  5. Is NTT’s AI inference chip available for commercial use?
    NTT’s AI inference chip is currently in development and testing phases, with plans for commercial availability in the near future. Stay tuned for more updates on when this groundbreaking technology will be available for use in various industries.

Source link