YEREVAN, FEBRUARY 17, ARTSAKHPRESS. Chatbots and artificial intelligence are modernizing the furniture industry, said Pavel Dimov, Board member of the Directors of the Branch Chamber of the Woodworking and Furniture Industry and managing partner of MOS Robotics, quoted in a press release by the organization. Dimov was adamant that artificial intelligence has its place in the furniture industry and it is important for companies in the sector to invest in analysis and processing of data obtained from production.
Dimov explained that artificial intelligence is simply the modern term for software algorithms that basically process large amounts of data. As in any industry, the equipment in the furniture sector is constantly digitized and automated, collecting operational data. Providers of services, software solutions and equipment strive to continuously improve customer performance by using and analyzing this data. Any company that wants to increase its efficiency should invest time and money in analyzing its processes, that’s where artificial intelligence can be extremely useful and save time for manual data processing, the expert pointed out.
Dimov noted that Industry 4.0 platforms, technique and software solutions reduce the prices of technology and allow any type of equipment to be continuously connected, online and available for analysis and automation.
Manufacturers such as HOMAG, KUKA Robotics, Universal Robots have been providing connectivity options for more than 15 years, and in recent years this has also been possible through cloud systems for remote access and analysis, as well as eliminating the need for servers and multiple expensive systems in any furniture factory.
Chatbots are simply a way to communicate with the end user, argued Pavel Dimov. AI processes an extremely large amount of data, and by implementing systems such as chatbots, out of all data on the servers the user gets only the parts, services, documentation, drawings, etc. that are important to them.
The technology that ChatGPT uses is not innovative in itself, said Dimov, this kind of Machine Learning models have been known for several years. What makes ChatGPT so good and distinctive is the huge scale of data (about 45 terabytes) it is trained on and the correspondingly expensive resources (servers) that are used for the purpose.
ChatGPT training costs several million dollars, the expert pointed out. ChatGPT is very interesting and useful for the average user, because it can write an article on a given topic, answer any questions, etc. However, from a business perspective, the problem is that ChatGPT makes mistakes because it is trained on a huge amount of internet data that includes all kinds of opinions and claims, even contradictory ones. A business cannot afford such mistakes, Dimov pointed out. It is therefore important for companies to focus on personalized business solutions that are trained only on what the customer requires, so as to avoid the danger of a completely uncontrollable response, as is the danger for example with ChatGPT.
When developing such type of chatbots for business purposes, usually the companies that provide the service use their own Machine Learning technology, which trains only data that is important to the customer, with specific terms.
Chatbots for business purposes can respond much more comprehensively than just text like ChatGPT. They also have a user interface that includes buttons, videos, images, web forms, etc. In addition, similar technologies created especially for a given company work with over 100 languages, including Bulgarian, Pavel Dimov pointed out.