In today's world, we face complex and dire challenges. The only certainty is change. Predicting future developments for the benefit of all requires collecting and analyzing information on an unprecedented scale. Artificial intelligence (AI) will play a vital role.
answer? Innovative Optical and Wireless Networks (IOWN). IOWN is a communications infrastructure that uses optical and photonic technologies to deliver ultra-high capacity, ultra-low latency, and ultra-low power communications.
Expectations and demands for AI are highly relevant to IOWN. Consider asynchronous distributed learning, where AI systems from different domains collaborate and share knowledge.
This helps solve the problem of centralized data collection. Centralized data collection is an impractical approach for diverse data sources, including vehicle, factory, personal, environmental, and countless other sensors.
IOWN serves as the communications infrastructure for such systems, allowing AI to process vast amounts of information and enhance interactions.
What is a large-scale language model (LLM)?
Large-scale language models (LLMs) are a rapidly evolving field in the field of natural language processing that enables AI to understand and generate everyday human language.
Progress will depend on expanding data availability, increasing computational power, and developing new training algorithms.
Some early LLMs have already begun to be widely adopted and are expected to have a significant impact on business and society as a whole.
For example, NTT has been engaged in research and development of natural language processing technology for many years.
Excited by the potential that AI has in improving the well-being of people around the world, the company launched its own LLM called tsusum. Tsusum was developed with an energy-efficient design, language processing capabilities, and adaptability to different user needs.
Rapid advances have enabled LLMs to interact naturally with humans. However, it is not without ethical and technical challenges. For example, LLM is susceptible to learning bias from training data, which can result in incorrect output.
Despite the great capabilities of LLMs, seamlessly collaborating with humans remains difficult.
Additionally, the internal workings of the LLM are not yet fully understood, making it difficult to understand how the LLM produces its output. Therefore, further research and development is still needed and will continue for some time to come.
The future of NTT and AI
NTT's goal is to develop AI cognitive engines that can naturally collaborate with people and contribute to society and individual well-being.
This means developing AI that has the same interface as humans. NTT is currently developing “VisualMRC,'' a model that visually interprets language in web pages in a human-like manner, and “SlideVQA,'' a model that responds to questions based on multiple images such as slide shows. NTT is also building a visual reading model for Japanese.
With these models in place, NTT will be able to create versatile software robots that can interact and collaborate with humans.
The idea is that everyone will be able to collaborate with this software as an assistant, and communication will increase not only between humans and AI, but also between AI and AI, and even between AI and objects around the world.
IOWN helps make all these advances possible. Imagine the huge amount of data that needs to be processed in real time, including text and all the audiovisual information that humans perceive.
IOWN enables the connection of these massive amounts of data generated by people, devices, sensors, and the digital world, enabling collaboration between humans and cutting-edge AI technologies.
Learn more about IOWN here.