Global Courant 2023-04-17 05:00:19
Global Courant: Do Japanese manufacturers use ChatGPT?
ChatGPT: It is possible that some Japanese manufacturers use ChatGPT or other similar language models for different applications… However, I do not have access to specific information about which Japanese manufacturers currently use ChatGPT or other similar language models…
Global Courant: what about Panasonic?
ChatGPT: I do not have access to information about whether Panasonic specifically uses ChatGPT or other similar language models. However, it is possible…
Global Courant: What about this?
Nikkei Asia: Panasonic unit deploys ChatGPT-like AI to improve productivity (March 8, 2023) OSAKA/PALO ALTO, USA – A systems development company under Panasonic Holdings has begun large-scale deployment of easy-to-use, interactive artificial intelligence that it hopes will improve productivity.
Panasonic Connect has made it possible for all of its 12,500 employees in Japan to use the AI as an “assistant” for tasks such as drafting documents, Nikkei has learned.
Through a collaboration with Microsoft, the company has developed an AI assistant that answers employee questions, using technology from OpenAI, an American company that developed ChatGPT, an interactive AI….
The company uses the AI system to produce statistics in a format that can be easily processed by spreadsheet software, create templates for documents, and solicit ideas on how to proceed with meetings….
The system can only be used within the Panasonic Connect internal network to prevent information leakage. To mitigate such risk, Panasonic Connect has agreed with Microsoft not to use content for secondary purposes such as AI learning. In addition, the company has warned employees not to enter any personal or confidential information.
ChatGPT: Thanks for the update! It appears that Panasonic Connect, a systems development company under Panasonic Holdings, has indeed deployed an AI assistant that uses technology from OpenAI, the developer of ChatGPT, to improve productivity, according to the Nikkei Asia article….
Look after
Watch out, shorter viewfinders. The previous conversation shows how the verbose ChatGPT, even when asked the name Panasonic, failed to notice an article published five weeks ago on a well-known English-language business website. Google found the article in less than a second.
And from the article, we learn that – before Elon Musk, Steve Wozniak and more than a thousand other individuals involved issued a “call” on March 14 for all artificial intelligence labs to “immediately begin training more powerful AI systems for at least six months.” interrupt” than GPT-4″ due to “serious risks to society and humanity” – Panasonic had already started implementing a low-risk internal version of the technology.
So we’ve been caught up in the furore over ChatGPT, the most publicized information on the subject of AI in Japan, but in the process, that furore has diverted us from more important applications across a range of industries. Examples include:
Pharmaceuticals: Astellas Pharma has created a cellular drug discovery platform that integrates humans, AI and robotics that is claimed to reduce the time to acquire drug candidates by approximately 70% in successful cases. It is now being used in a joint research project with Kyoto University’s Center for iPS Cell Research and Application to “advance the use of differentiated cells and tissues derived from human-induced pluripotent stem cells (iPS cells) and to develop innovative creating medical solutions. Regenerative Medicine: Canon plans to apply its deep-learning neural network for image processing to regenerative medicine following the acquisition of Kyoto Seisakusho’s high-speed cell culture system technology (completed in May). Factory worker support: Hitachi has developed a “real-time work anomaly detection” system based on deep learning that reduces inefficiencies caused by human-machine interaction in highly automated factories. The company also uses neural networks to improve fault detection in complex industrial and infrastructure equipment, systems and networks. Once an industrial hardware conglomerate, Hitachi now applies digital technology to everything it does. Supply chain logistics: Hitachi has also developed algorithms to maximize the efficiency of complex supply chain networks where multiple factories, warehouses and stores are connected by land, sea and air. This includes production planning, inventory management and maximizing the efficiency of the delivery route. Quality Control: NEC uses machine learning to standardize and accelerate product inspections, saving time once spent on visual inspection and improving accuracy through objective data analysis. NEC also uses automated real-time analysis of temperature, vibration, pressure, voltage, flow rate and other data from sensors to monitor power generation facilities. This “allows the detection of anomalies at an early symptom stage, including subtle anomalies that would go undetected even by inspection experts.” Industrial Cybersecurity: Fujitsu provides services ranging from the assessment of a manufacturer’s operational technology to the design and implementation of cybersecurity systems, continuous monitoring, incident investigation and system upgrades. They are based on data analysis and stem from Fujitsu’s own experience as a manufacturer of computer and communications equipment. Embedded Systems: Semiconductor maker Renesas has put artificial intelligence on embedded systems (e-AI) to run inferences on its MCUs, MPUs and AI accelerators. This enables real-time processing without the network lag inherent in cloud computing. Applications include factory automation monitoring (for example, to catch abnormal vibrations), automated system device coordination and motor control or other failure prediction. Semiconductor metrology: Hitachi applies deep learning to critical dimension scanning electron microscopy (CD-SEM) to more accurately measure defects in high aspect ratio holes in 3D semiconductor devices (e.g. NAND flash memory) after etching to improve the process check. Hitachi also makes etching equipment.
The list goes on and on, from healthcare, traffic control, waste management, wind turbine performance – anything that requires huge amounts of data to be processed. All this is way beyond ChatGPT, although that doesn’t mean it couldn’t take your job.
Definitions
IBM points out that “Perhaps the easiest way to think about “artificial intelligence, machine learning, neural networks – which mimic the human brain through the use of algorithms – and deep learning” is to think of them as Russian nesting dolls. is essentially a part of the previous term.
In other words: “machine learning is a subfield of artificial intelligence. Deep learning is a subfield of machine learning and neural networks are the backbone of deep learning algorithms.”
Arm, the IC design company, tells us that:
AI Inference is achieved through an “inference engine” that applies logic rules to the knowledge base to evaluate and analyze new information. There are two stages in the machine learning process.
First, there is the training phase where intelligence is developed by capturing, storing and labeling information. For example, if you train a machine to identify cars, the machine learning algorithm is fed with many images of different cars that the machine can later refer to.
Second, there is the inference phase where the machine uses the information collected and stored in phase one to make sense of new data. At this stage, the machine can use inference to identify new images and categorize them as “cars”, despite never having seen them before. In more complex scenarios, this inference learning can be used to improve human decision making.
Similarly, if you send a news article to ChatGPT, it will quote it directly to you a moment later. But it will not lead the industrial economy.
Follow this writer on Twitter: @ScottFo83517667
Similar:
Loading…
