What Can AI Technology Do for the 'Psychological Pandemic'
-
Under the shadow of the pandemic, we must not only fight the virus but also pay attention to the psychological well-being of the public. Confronted with the mental stress brought by COVID-19, AI technology is assuming an increasingly significant role in psychological care.
Over the past two weeks of the 'nationwide epidemic prevention' efforts against COVID-19, not only has the psychological intervention for frontline medical workers garnered widespread attention, but the general public also seems to have reached a critical point.
From the lighthearted 'creative broadcasts' and online surveillance as ways to pass the time, to the eruption of public outrage over certain incidents and the continuous rise in case numbers, people have collectively realized that this is not a battle that can be 'cute-fied' or swiftly resolved. Instead, it requires sustained combat, relentless efforts, and prolonged hardship.
Whether it's the frontline 'warriors' working long hours away from their families or the anxious general public confined at home, beyond practical measures like information transparency, material support, and safety precautions, there is an urgent need for psychological healing and redemption.
Today, let's explore what AI can do to address the 'psychological pandemic.'
This nationwide mobilization against the epidemic has swept people from all walks of life into its fold, directly leading to the diversity and severity of psychological issues. While we cannot exhaustively list all 'psychological ailments,' here’s a broad, albeit not entirely rigorous, categorization:
Frontline medical workers, who courageously shoulder heavy workloads, are also the group most exposed to danger and psychological trauma. Separation from family, the challenges of treatment, the helplessness in the face of death, the breakdown from overwork, and the anxiety of infection—all these complex emotions pile up, forming a composite stressor that affects them physiologically, psychologically, and socially, leaving them in a state of high psychological stress.
For instance, physiologically, some doctors in Wuhan have said they 'cried a month’s worth of tears.' Professionally, they may also struggle with feelings of inadequacy, doubting their ability to save patients and questioning their professional worth, leading to excessive self-blame. The Shanghai Public Health Clinical Center’s support team, for example, includes seven psychotherapists.
Clearly, the physical and mental exhaustion of medical workers not only pains the public they protect but may also impact patient safety and healthcare quality. Beyond ensuring their work safety and physical health, addressing their feelings of isolation and helplessness is also an urgent priority.
Another group under immense pressure includes those with suspected symptoms like fever and confirmed patients. Especially right after hospitalization, they may experience intense negative emotions—panic, worry, anxiety, suspicion—and even exhibit uncooperative or harmful behavior toward medical staff.
Dr. Zhang Wenhong, head of the Shanghai Medical Treatment Expert Group and director of the Infectious Diseases Department at Huashan Hospital, noted, 'Sometimes, patients don’t die from the disease itself but from fear.'
Currently, the root causes of these negative emotions can be traced to three main aspects:
Beyond criminal penalties for a few malicious incidents, the key to improving doctor-patient relationships, fostering cooperation, and winning the battle lies in identifying and providing timely psychological intervention and counseling.
Standing alongside medical workers and patients are the hundreds of millions of ordinary people who have voluntarily self-isolated and reduced outings due to the epidemic. However, as the turning point in epidemic control remains elusive, perceptions of the disease’s contagiousness and severity can easily become 'overcorrected,' leading to hypochondria—where a single cough might prompt someone to 'turn themselves in' or trigger a rush on medical resources and the spread of misinformation, inadvertently putting themselves at greater risk.
At the same time, constant attention to pandemic updates can easily lead to emotional volatility and 'compassion fatigue.' Netizens in recent days have likely experienced the rollercoaster of emotions from 'fuming with anger' to 'sobbing with感动555.' Beyond anger and heightened sensitivity, excessive exposure to negative news can also lead to burnout and desensitization, making it harder to empathize with other information.
After the pandemic is under control and life returns to normal, lingering psychological aftereffects may include rumination over events during the outbreak and fear of crowded places.
Of course, we must acknowledge three realities that make psychological intervention at this stage no easy task.
In fact, psychological services for medical workers, patients, and the public have been rolled out alongside epidemic control measures.
The National Health Commission has mandated that local authorities establish psychological assistance hotlines for COVID-19 under the joint prevention and control mechanism, with each hotline operating at least two lines to provide 24/7 free psychological services tailored to local needs.
Many experts and volunteers in psychotherapy have also stepped up to offer online counseling. However, addressing the psychological stress of such a massive population remains a drop in the bucket.
Frontline medical workers and patients are often in closed environments, and many counseling volunteers fear infection, leaving only a fraction of psychological resources available for direct intervention. Phone-based counseling, meanwhile, cannot capture subtle real-time reactions, potentially adding to the stress.
Even when psychologists can engage, verifying information from various sources, understanding the personal circumstances of those they’re speaking with, and grasping disease-specific knowledge can pose challenges. Coupled with the fact that China’s public mental health mechanisms have never faced a challenge of this scale, helping more mental health professionals quickly adapt and acquire relevant skills has become another hurdle in delivering psychological services.
This is where AI might step in as a useful 'assistant.'
Can AI deliver 'humanistic care'? We believe AI can contribute in three key ways:
First, AI can conduct 'preliminary screening' for patients seeking care. Using infrared recognition, machine vision, and big data models, combined with vital signs like body temperature, CT scans, blood pressure, and blood oxygen levels, AI can enable rapid triage, reducing the diagnostic burden on doctors and minimizing potential friction between triage nurses and emotionally unstable patients.
For example, Washington State once deployed an AI system in emergency rooms, which reduced ER visits by 10% by providing doctors with pre-screening information.
Second, for doctors working long hours in protective gear, communication often relies on walkie-talkies, and even gestures like elbow bumps for encouragement. Some doctors also feel that expressing helplessness might seem 'weak' or 'making a fuss.' Could AI add a touch of care to their work environment?
The AI robot Cimon, sent to the space station by SpaceX, offers an interesting example. Though its round face and enigmatic 'hehe'-like smile drew online ridicule for being 'too creepy,' in the confined, lonely space environment, Cimon provided companionship through facial expressions and listened to astronauts’ feelings. Today, with voice technology from smart assistant manufacturers already mature, introducing such systems into hospital wards could offer doctors a sympathetic ear.
Patients with symptoms, whether in beds or waiting in hospital halls, often face anxiety. With treatment resources stretched thin, doctors have less time for detailed communication. Here, AI’s service capabilities can be effectively activated.
Beyond conversation, AI can gather patient information, analyze preferences, work status, and medical needs using backend data, and build psychological profiles to deliver sustained, targeted communication and counseling, alleviating loneliness and anxiety.
In late 2017, Rohit Prasad, chief scientist for Alexa, noted that every interaction with Alexa-enabled devices collects voice recordings to analyze users’ emotional states. This enhances Alexa’s emotional recognition, paving the way for Amazon’s robots to comfort patients and provide warm companionship in hospital rooms.
In artist Dan Chen’s 'Life Care Machine,' a robot tells a bedridden patient, 'I’m sorry your family and friends couldn’t be here. I’ll do my best to take care of you.' Imagine the comfort such an emotional machine could bring to the isolating reality of quarantine.
Beyond the previously mentioned smart chat (I've personally engaged in 800 rounds of idiom battles with Siri), what I really want to highlight is the Text-to-Speech (TTS) technology.
We know that in this battle, hundreds have lost their lives, meaning hundreds of families are plunged into grief and loss, with some even unable to bid a proper farewell to their departed loved ones. Heartbroken as we are, what can AI do for these 'survivors'?
Preserving their 'voice and smile' might offer some solace. Previously, a foreign journalist, James Vlahos, created a chatbot simulating his late father's consciousness using his voice.
With 'small data generation technology,' creating a custom voice package doesn’t require much data—just 20 sentences can suffice. By decomposing and processing the recordings, machine learning can synthesize a simulated voice. This allows the voice of the deceased to be recreated on devices like phones, computers, or car tablets, keeping their memory alive.
For instance, advancements in NLP technology itself are crucial. Emotion recognition must be combined with real-world dialogue scenarios to ensure accuracy. This requires multi-dimensional data collection and analysis, establishing analytical perspectives tailored to different service subjects, adjusting parts of speech and sensitivity, and matching the emotional weight of various words to build more accurate models. There’s still a way to go before it can be fully operational.
Moreover, given the involvement of personal privacy and critical medical information, the security of such AI services must be highly prioritized.
In summary, AI itself has no emotions. The humanistic care it exhibits comes from the employees, companies, and even the government and society behind this field. Hopefully, in the future, we’ll see AI not only help break through the defenses of disease but also assist all those involved in rebuilding their spirits and lives. This is the ultimate blessing of technology.