【推荐1】Landscape architect Kotchakorn Voraakhom has designed a new green roof on the Rangsit Campus of Thammasat University, about 25 miles north of central Bangkok, Thailand. Her imaginative work challenges the common thinking that urbanization has a negative impact on the planet, whether flooding, excess (过度的) energy use, disrupted (扰乱) biodiversity or the heat island effect.

The 236, 806-square-foot structure, which opened in December 2019, includes a flood water management system and Asia’s largest rooftop organic farm. It combines modern landscape architecture with traditional agricultural knowledge, creating a green and friendly environment.

The green roof, containing an H-shaped landscape, looks like a futuristic hill with a brick building beneath it. The hill features a complex pattern of zigzagging terraces (之字形梯田) of planted beds, leading all the way down to the bottom. When rainwater hits the roof, it flows down the zigzags while being absorbed by the soil in the beds, The excess water is directed into four storage ponds — with a capacity of up to 3 million gallons. The process slows down the flow speed of rainwater runoff compared to a normal concrete rooftop. This keeps the area from flooding during heavy rains.

The roof’s terraces are filled with organically grown crops, including a drought tolerant variety of rice, many local vegetables and herbs. The farm can supply the canteens on campus with a large amount of rice, herbs and vegetables a year. The food waste is composted (把……制成堆肥) to fertilize the farm, and water from the storage ponds is used to water plants, creating an entirely localized and circular system.

The farm serves as an outdoor classroom and a source of local jobs, too. Farmers offer workshops on sustainable agriculture and nutrition as part of the university’s sustainability curriculum. “Students and community members are invited to participate in seasonal seeding, harvesting, and so on,” says Voraakhom. “The urban farm is training a new generation of organic farmers with real-world skills. It also promotes a sense of community.”

【推荐1】AI may sound like a cold robotic system, but scientists from Osaka Metropolitan University have shown that it can deliver heart warning, or more to the point, “heart-warning” support. They find a new use of AI, diagnosing (诊断) functions of our heart and valvular (瓣膜的) heart disease with its accuracy, which shows continued progress in combining the fields of medicine and technology to advance patient care.

Valvular heart disease, one cause of heart failure, is often diagnosed by echocardiography (心超声检查). This technique, however, requires specialized skills. However, there is a shortage of qualified technicians. Meanwhile, chest X-ray is one of the most common tests to identify diseases, mainly of the lungs, and very little time is required to conduct them, making them highly accessible. Even though the heart is also visible in chest X-ray, little was known about their ability to identify heart function or disease. Therefore, the research team led by Dr. Daiju Ueda from Osaka Metropolitan University, thought that if heart function and disease could be determined by chest X-ray, this could serve as a supplement to echocardiography.

To build the relationship between them, the team turned to the AI model. They researched as many related data as possible. A total of 22,551 chest radiographs associated with 22,551 echocardiograms were collected from 16,946 patients between 2013 and 2021. With the chest radiographs set as input data, and the echocardiograms set as output data, the AI model was trained to learn features connecting both data.

The AI model was proved to be able to identify exactly the types of valvular heart disease. “It took us a very long time to get to the results, but I believe this is a significant research. In addition to improving the efficiency of doctors’ diagnosis, the system might also be used in areas where there are no specialists, in night-time emergencies, and for patients who have difficulty undergoing echocardiography,” Daiju Ueda said.

【推荐2】Is it time to put the brakes on the development of artificial intelligence? If you've quietly asked yourself that question, you're not alone. In late March, a group of AI experts signed an open letter calling for a six-month pause on the development of more powerful models than GPT-4; European researchers called for tighter AI   regulations; and long-time AI researcher Yudkowsky demanded a complete shutdown of AI development.

Meanwhile, the industry shows no sign of slowing down. In March, a senior AI executive at Microsoft spoke of “very, very high”   pressure from chief executive to get GPT-4 and other new models to the public   “at a very high speed ”. GPT-4 is much larger and has been trained on significantly more data. Like other large language models, GPT-4 works by guessing the next word in response to prompts(提示). In tests, it passed legal and medical exams, and can write software better than professionals in many cases. And its full range of abilities is yet to be discovered.

GPT-4 and models like it are likely to have huge effects across many layers of society. They could facilitate personalized phishing(网络钓鱼) attacks, produce disinformation at scale, and be used to hack through the net -work security around computer systems. Open   AI's own research suggests models like GPT-4 are   “general-purpose technologies” which will influence 80 percent of the US workforce.   Professions such as customer services, primary translations and editing, and security guard will be greatly impacted. Moreover, technology is accelerating much faster than our ability to understand and regulate it. If we're not careful, it will also drive changes that are too fast for safety. The US sociologist E. O. Wilson described the dangers of change like so:   “The real problem of humanity is the following: we have   Paleolithic(旧石器时代的) emotions, medieval(中世纪的) institutions and god-like technology.”

I believe a wise course of action is to slow down and think about where we want to take these technologies. It is not about stopping, but rather moving at a sustainable pace of progress. We can choose to steer this technology, rather than assume it has a life of its own we can't control.

【推荐3】An AI model can be used to detect stress in office workers based on how they use their mouse and keyboard. A new study suggests that a machine-learning model using these two elements was more accurate at detecting stress in people than a model that tracked their heart-rate data.

“We saw that the models that just used the mouse and keyboard data performed better than the models that had the heart-rate data in it,” says Mart Naegelin. a Ph. D. student at the Swiss Federal Institute of Technology, and one of the study’s authors.

Naegelin and her fellow researchers used machine-learning models to analyze data on keyboard-typing activity, mouse movements and heart-rate data. They also studied the data of a combination of two or all three of these elements to determine which performed best in terms of measuring stress. They found that the model trained on mouse and keyboard data performed better than the model that used mouse, keyboard and heart rates. Of the single-modality (形态) models, the heart-rate data performed the worst. “The test was conducted in an environment that simulated an office environment, so the results still need to be confirmed in real-life scenarios (情况),” Naegelin says.

In the experiment, participants were divided into three groups. A control group carried out assigned tasks, such as planning meetings and collecting data, with no additional work. A second group at times answered questions from managers in person in a mock interview scenario while completing other assigned tasks. And members of a third group were, at certain points, interrupted with additional questions sent through an online chat on top of tasks assigned to other groups. At regular intervals during the experiment, participants were asked to rate their stress levels through a computer questionnaire.

Researchers learned that workers made longer, less accurate movements with their mouse, as well as more typing errors, when they were stressed. Shorter, more direct movements with the mouse were tied to lower stress levels. The study didn’t focus on why increased levels of stress are thought to affect muscle activity.

The researchers say they believe a stress-detection system that logs keyboard and mouse movements might be beneficial as a self-help tool for employees alongside other initiatives to improve workplace mental health. But employee participation would need to be optional and based on informed consent (同意), and companies would need to commit to protecting user privacy.

It remains to be seen how this technology will evolve and be adopted in practice, but the potential for AI to contribute positively to employee well-being is an appealing avenue for future exploration.