Leading companies in construction, travel, retail, healthcare, and energy production technologies appear to have recognized the importance of AI in improving productivity and modifying consumer behavior. However, at the same time, these companies face difficulty in bearing the high expenses of this technology, and the weaknesses that still accompany it.
Author Bill Lane says in a report published by the American newspaper “The Wall Street Journal” that generative artificial intelligence received great attention during this year, as it is a pioneering technology in various sectors of the economy that promises us to raise productivity to record levels, by performing many jobs and improving work methods. .
But due to the high cost of artificial intelligence, the constant need for specialized engineers in this field, in addition to the legal risks and the dilemma of respecting personal data, the progress that has been made in this field is still limited, in light of the caution that has overwhelmed the behavior of many institutions that are trying these new solutions. .
But despite this hesitation, there is no doubt that generative AI will completely change the landscape of economic technologies. Some companies have begun to include artificial intelligence in their operations to write codes and programming, create marketing content, and provide customer support.
By the end of the year 2023, various companies around the world spent $19.4 billion on generative artificial intelligence, and this amount, which includes generative artificial intelligence software and accompanying equipment and other services, will jump to $151.1 billion by the year 2027, i.e. an annual growth rate of 86.1% over the course of 2023. four years.
The author explained that most companies are currently testing the limits of this technology, and increasing investment in artificial intelligence and cloud computing more broadly in organizing their databases and assessing cybersecurity.
Many experts and pioneers in this field are looking forward to the year 2024, as it will be the year in which generative artificial intelligence will prove that it is important and worthy of the huge sums spent on it.
The author reviews five companies that were able to find solutions to integrate generative artificial intelligence into their products. Among these companies are the travel agency Expedia, the health care system Mass General Brigham, and the construction business software Bentley Systems.
Wayfair.. Artificial Intelligence, the new interior designer
The writer indicated that this company, which sells furniture online and is headquartered in Boston, USA, uses generative artificial intelligence to help customers redesign living rooms.
All it takes is for the customer to upload a photo of his room for Wayfair software to use generative artificial intelligence models to produce multiple versions and styles of what the living room could look like after purchasing new furniture.
Along with the design tool, Wayfair has generative AI applications that improve productivity and have proven economic feasibility.
But according to the company's CTO, Fiona Tan, in some operations, Wayfair intends to continue to rely on machine learning models that are less expensive, can produce improved perceptions, predictions, and solutions, and are therefore more effective than generative artificial intelligence.
The author points out that the interior design and decorating tool provided by Wayfair is available to users for free, and it produces very realistic images. But since it relies on generative artificial intelligence, it occasionally produces images that contain strange errors and defects, such as a table with unnatural legs, or as if the artificial intelligence is sometimes unable to determine the location of windows and a mirror.
Schneider Electric.. Artificial intelligence to improve energy consumption
According to the author, this company is considered one of the largest manufacturers of electrical and automatic products, and it prefers to use smaller, less power-consuming artificial intelligence models in some limited scopes.
The company's Artificial Intelligence Director, Philip Rambach, says: “Large language models can summarize data and produce texts, but when it comes to predictions and improving processes, they are not effective enough, and they drain a lot of money and energy, so their use has not yet become a decision.” “Wise.”
The author points out that generative artificial intelligence drains huge amounts of energy from data processing centers, relying on electrical energy, which has become more expensive to produce. Artificial intelligence could consume up to 3.5% of the world’s electricity by 2030, according to a study completed by technology consulting and research firm Gartner.
That's why Schneider Electric, which produces home electrical appliances, electric car chargers, home automation systems and energy management software, chose to rely on the GPT 3.5 chat model, which is smaller and much less power consuming.
This company, based in France, also uses generative artificial intelligence to help its customers calculate and analyze the gas emissions resulting from their energy consumption.
Brigham's Mass.. Data-Driven Screening
The writer mentioned that this company, which is based in Boston and is associated with Harvard Medical School, uses generative artificial intelligence to identify patients who have common characteristics, and thus assist doctors with this information in order to treat the patient based on successful experiences in similar cases.
The company's healthcare system relies on the artificial intelligence medical imaging company Analyze AI, along with other startups, to bring new solutions to this sector. The company is currently preparing to launch the first medical diagnostic products using artificial intelligence during the first half of this year.
According to the company's director of data science and deputy director of medical imaging, Keith Dreyer, this scientific field is advancing at a very rapid pace. Drawing on pre-existing data, medical files, conversations with patients and genetic characteristics, large language models can identify patients who are similar to each other.
This type of technological solution in the past required seven years of training and millions of dollars to develop artificial intelligence algorithms capable of analyzing images, but now large language models can be trained within a few months.
Although this solution remains very expensive, it is fast, and therefore in the long run it will save costs, especially since it is able to find points of similarity between texts and x-ray images.
Expedia..your personal assistant for booking flights
The writer reported that Expedia relies on chatbots to help travelers by providing advice, recommendations, and technical assistance when making reservations.
However, this company does not consider artificial intelligence a magic solution, as Expedia CEO Peter Kern says: “Artificial intelligence is just an assistant, and it will enable us to achieve our goals and explore some new solutions, but the idea of asking artificial intelligence to book an entire trip with all its details is an issue that involves… Exaggeration. Instead, Expedia, which is headquartered in Seattle, will improve users' experience while booking their flights. For example, the company's website will be able to answer customers' questions based on their previous travel history and the things they prefer, and thus the answers will be more personalized. And realistic.”
Bentley Systems: more solid infrastructure and automated designs
The writer mentioned that this company, which is based in Extension, Pennsylvania, produces software for the building and construction sector, and is now developing artificial intelligence tools capable of formulating drawings and designs for project sites, and the forms of infrastructure and buildings in a more responsive manner to the challenges of environmental factors and sustainability.
“We are asking the question of the impact and impact relationship between the Earth's temperatures and the highways, bridges and other infrastructure,” says the company's technology director, Julian Mott. “We must be able to analyze designs to assess future conditions so that our projects are able to withstand natural conditions.” difficult.”
In the short term, Bentley Systems is working to help engineers produce designs using artificial intelligence in order to provide an automated assistant that will participate with engineers in 30 to 50% of this work.
The author concluded the report by saying that artificial intelligence has already begun to be used in the field of infrastructure to treat cracks and rust in bridges, and to predict when roads will need maintenance and restoration work.
In addition, the large language models now used by Bentley Systems are capable of analyzing huge amounts of data, including analyzes of the condition of structures and facilities, building codes and laws of physics, climatic conditions, and the effects of new building materials on future designs.