Recently, there isn't a day that I don't encounter the term “generative AI” represented by chat GPT. The degree of innovation of this technology from the United States, which was unveiled in November last year, can be estimated by the degree of penetration into society in such a short period of time.
Regarding its utilization, it seems that there are various high hurdles depending on the industry and type of business, but I think it is necessary to challenge it from the viewpoint of improving the sophistication and efficiency of operations.
In other words, since generative AI instantaneously summarizes documents and prepares planning proposals (even if there are mistakes), productivity is improved, and surplus forces cannot be replaced by AI, for example, can be diverted to enhancement of sales and consulting departments. However, what should be kept in mind is that it is not enough to “throw a round throw” at AI, and it is necessary for humans to find mistakes, train AI by changing the way questions, etc., and make efforts to bring it closer to ideal results. “Training” means accumulating learning, where control based on human wisdom and experience is necessary, and in the end, human abilities are questioned. It is also directly linked to a company's ability to develop human resources.
In the real estate industry, the quality of property guidance created by AI generated in the distribution market is good, and it seems that they are solidifying their ground in that field, so it can be said that it is a major achievement. According to recent reports, it is said that the government will begin operating a database of “real estate IDs,” which assigns identification numbers to each real estate such as land and buildings, in some municipalities by the end of the year, so it will be possible to obtain construction regulations, hazard information, and gas and electricity usage information all at once. I don't think all of the information necessary when brokering or appraising real estate can be obtained (construction confirmation information that has expired at government offices, etc.), but once such data infrastructure is developed, it will also be a major support material for the advancement of generative AI.
There are currently many negative opinions about the use of generative AI in real estate appraisal sites. This is because field surveys and government office surveys are essential, and compatibility with appraisal work, where judgment work starts from there, is not visible. However, instead of expecting too much from generative AI, I think the finished form will come into view by repeatedly questioning the draft contents created by AI on the basis of various information that real estate appraisers have conducted on-site investigations. Projects that are similar in type to be addressed next can be completed without spending that much time. In real estate appraisals, unlike requesting summaries and translations from generated AI, it takes a considerable amount of time and effort to reach the finished form. However, even so, if we master AI and refine our skills, we should be able to blow a new wind into conservative industries.
Reprinted from “Weekly Building Management” by Building Management Institute Co., Ltd. (with permission)
