As machines get smarter, we’re using them to help us with more and more. Everything from smartphone assistants like Siri, to chess-playing robots, to self-driving cars, operates on a technology known as artificial intelligence or AI.
While AI often brings to mind frightening dystopian scenarios in which intelligent robots take over the world, today they’re allowing us to accomplish a wide range of tasks more efficiently and successfully. The technology is still in the relatively early stages though, and there’s a lot we still don’t know about how it works, what it’s capable of and how best to use it.
Healthcare is one area where we recently started using AI. Intelligent machines have shown a lot of potential for helping us with health-related issues. It’s been used in medical imaging, to provide mental health tips and to help people overcome addiction. Some experts warn, however, about relying too much on intelligent machines, especially in this early stage of their development.
About Artificial Intelligence
Artificial intelligence is a type of computer science that seeks to make machines “intelligent” in a manner similar to that of humans. Intelligent machines can learn and make decisions without having to be told explicitly what to do by a human operator at all times.
In order to make intelligent decisions, machines need access to sufficient information through a database or other means. Giving them this access is known as knowledge engineering. They may also gain access to input through machine perception, which includes things like computer vision through the use of a camera. Equipping machines with vision and other senses allows them to gain information from the world around them.
Smart machines also need to be able to interpret this data and identify patterns within it. Then, they must use these observations to draw conclusions and make decisions. This ability is a central part of AI and is known as machine learning. Intelligent machines can sift through given information and even find their own information. In general, the better a machine is at interpreting information, the smarter it is.
Other modern technologies also intersect with AI. Making machines more intelligent is crucial to robotics, for instance. To be more effective, robots need to know how to interact with new situations safely and efficiently. The Internet of Things (IoT) is another new technology that influences the development of AI.
This network of devices uses the Internet to connect to each other and share data with each other to coordinate their operation. For example, a smart porch light would turn on when it senses that your smart car is pulling into the driveway. These devices often use machine learning to improve their performance over time. The porch light might learn when you typically come and go, for instance. These smart devices collect huge amounts of data, which they can share with other devices, allowing them to get smarter faster.
Today’s AI falls under the category of narrow or weak AI. This classification means that it’s programmed to do just one specific task, such as play a game of chess or drive a car. Many researchers are working toward general or strong AI, which would be a more complete form of intelligent machine. General AI could theoretically learn to do any task.
Some researchers think that if we give machines enough data and combine all the databases used by AI entities, machines could eventually reach an intelligence similar to that of human beings. Many others believe that we need other advances in the field of AI before that could be achieved.
It isn’t clear exactly where AI is headed or when we’ll get there, but that hasn’t stopped us from trying out the technology we do have in the real world. Companies like Amazon use it to suggest products that you might like to purchase. Netflix uses an AI-powered algorithm to recommend shows and movies. Smart home devices like thermostats and light bulbs use it to learn your habits and optimize their operation accordingly. Even infrastructure like the electric grid could use artificial intelligence to manage power generation and distribution, so that the grid operates most efficiently and for the lowest possible cost.
Businesses use AI-powered chatbots, AIs that allow you to message back and forth with a computer, to improve their customer service. You may have seen examples of this if you recently visited the websites of any major retailer.
These bots can answer your questions and sometimes even recommend items from the company’s online store. These computer programs use the information they get from their chats to get smarter and improve their conversation skills over time. In today’s iterations, if they’re faced with a question they don’t understand or know the answer to, they usually transfer you to a human customer service representative.
Artificial intelligence is playing a similar role in healthcare through the use of chatbots, as well as showing up in many other places in the healthcare industry.
How is Artificial Intelligence Used in Healthcare?
Healthcare companies have already started using AI, but its influence on the industry is likely just beginning to take shape. One survey found that while just under five percent of hospitals were using AI, 35 percent planned to start using it in the next two years and more than half said they would start within five. The industry has found ways to use it for record keeping, information security, assistance during surgery, prosthetic limbs, mental health treatment and more.
Researchers are using artificial intelligence to develop a new prosthetic hand that can “see” and identify the objects in front of it. The hand is equipped with a camera that provides it with what’s known as “computer vision.” It takes a picture of the object and adjusts its movement so that it can grip the item. The more the hand interacts with objects, the better it gets at choosing the right grip, by using principles of machine learning.
AI is also being used to assist in medical imaging. Once it’s fed a large number of images and taught what to look for, a machine learning program can detect patterns in the images. This ability could help medical personnel to interpret this imagery more accurately. The computer program acts as a second set of eyes and might pick up on things that people missed. The fact that it has access to such a large volume of images helps it to interpret the information it’s presented with more quickly.
Documentation is crucial in the medical field. It helps doctors to diagnose health problems, ensures that patients get the right medication and keeps the whole patient care system organized. Because of the high volume of information involved, it can also get complicated. Several healthcare organizations have begun incorporating AI into their documentation to make it more accurate and efficient. These programs analyze records and use their database of other medical documents to identify possible missing information or other discrepancies. While AIs can’t yet create and manage patient records completely on their own, this innovation has the potential to increase patient safety.
Healthcare companies are also using AI to protect this important patient information. These AI systems analyze the state of facility computer networks and, because they know what the network should look like, they can identify abnormalities. These irregularities can sometimes be caused by cyber-attacks and other threats, which the system can then alert IT staff to.
Artificial Intelligence and Mental Health
AI has also begun to play a role in the treatment of mental health conditions. A digital health company called Tiatros Inc. started using IBM’s AI program Watson to help soldiers with post-traumatic stress disorder (PTSD). Tiatros created a private social network for soldiers where they could talk with their peers. Each participant wrote and submitted three narratives each week and commented on the stories of other soldiers.
Watson analyzed these stories and interactions and evaluated the emotional components in them. It then recommended personalized cognitive behavioral therapy strategies based on its findings. According to Tiatros and IBM, the program has a 73 percent completion rate, whereas similar programs that don’t use AI typically have a mere 10 percent completion rate.
AI Healthcare Apps and Chatbots
The use of AI for mental health treatment has been gaining popularity recently. A number of artificial intelligence-powered automated conversational agents, more commonly known as chatbots, have recently come onto the scene.
These bots allow users to message back and forth with them about their mental health. Although they’re just computer programs, they’ve been programmed to have conversations about mental health and learn how to get better at this as they go. This ability enables them to simulate a conversation with a human being.
Woebot, one of the most popular mental health chatbots, operates in Facebook messenger. Within the messenger app, an endearing-looking cartoon robot welcomes you to have a short conversation with it every day. It asks you about your mood and recent life events. Based on your responses, it recommends resources to you, such as videos and articles, about mental health.
It also tracks your mood based on what you tell it and informs you about overall trends in your emotional well-being. Unlike humans, Woebot has a perfect memory. It regularly reminds you about your mental health and helps you identify patterns in your thoughts, feelings and actions. This aspect of Woebot is meant to get you more in touch with your feelings, so that you can manage them more successfully.
The chatbot was programmed by a behavioral cognitive therapist and uses the principles of CBT to guide its interactions and recommendations. Its creators don’t claim that it can heal deep-set trauma or cure severe psychological issues. They have said that it can reduce symptoms of depression in two weeks though, in a study coauthored by Woebot’s founder and published in the Journal of Medical Internet Research.
Chatbots have also been developed for helping people to manage addictions, compulsions and bad habits. These computer programs operate similarly to Woebot. They ask users daily about their progress and can also track activity, if the behavior the user wants to change involves visiting certain websites, such as online gambling sites, or limiting screen time in general. The program can also block these problem sites if prompted to.
Users also have the option to chat with an AI about their addictions. The bot will respond intelligently and suggest resources relevant to the conversation. This skill, in combination with increased awareness about their habits from the AI’s tracking capabilities, may help people better manage their addictions and compulsions.
Using chatbots for mental health is a relatively new idea and it’s already showing promise. Some mental health professionals have expressed doubts, though, about its effectiveness and even its safety.
Is AI Qualified to Give Health Advice?
With the exception of the study conducted by Woebot’s founder who’s also a clinical psychologist at Stanford University, there hasn’t been much research on whether chatbots can really help people with their mental health. The idea is so new that we don’t really know yet if we can trust a robot to give mental health advice. It may be awhile until we know exactly how well they work.
There is some doubt that at this stage in artificial intelligence technology, robots can actually tell what we’re feeling. You could argue that will never happen, since there’s something about understanding emotion that requires a human touch.
Although AI bots can closely simulate a conversation with a human being, they have never had the human experience and, so, can’t relate nearly as well to people as a human therapist. They also can’t understand and appropriately respond to as many statements as a person can – at least not yet. If you say something to a robot that it hasn’t learned yet, it may not respond in an insightful way. Some of these misunderstandings might be things that a therapist would call attention to and ask a patient to challenge.
The fact that chatbots might not always respond appropriately raises some concerns. They could miss out on important warning signs or even give their conversation partners feedback that could be harmful. If the program is fed the right information and learns properly, it shouldn’t do anything that would cause people too much worry. There are, though, some aspects of human interaction that chatbots lack.
Woebot’s creator is aware of its limitations. If after six weeks, your mood hasn’t improved, it suggests going to in-person therapy. Chatbots are regarded as something of a first step that people can use to evaluate their mental health and get some useful information, tools and recommendations. No one claims that they’re a replacement for therapy, but in some situations, they may be the best available option.
Therapy can be expensive and some people may not be prepared to pay for it, at least until they’ve learned more about their mental health. For people who live in isolated areas, getting to a doctor’s office regularly may be difficult. People are often wary of opening up to other people, even therapists. Chatbots eliminate this barrier. Using a chatbot is very convenient and may introduce more people to mental health strategies at least until, if needed, they can seek other help.
The Benefits of Human Connection in Recovery
Although using technology in therapy has its benefits, you miss out on certain beneficial aspects of mental health care if you lose the human aspect. Even just building a trusting relationship with a therapist who shows empathy and understanding can be beneficial. In fact, in a recent survey, 67 percent of respondents said they preferred meeting with their therapist in person, while only seven percent indicated that they’d rather meet online.
Other people can readily give you signs that they empathize with you, even through non-verbal reactions such as nodding. People give their therapist non-verbal cues as well, often without even realizing it, that could be important to the meaning of what they say.
With robots you don’t get this. Because talking with a chatbot is completely text-based, the conversation is void of non-verbal cues. While there hasn’t been much scientific research on whether discussing your worries with a bot has the same benefits as talking about them with a person, one can see why there are likely some differences.
Even if the person you connect with isn’t a mental health professional, confiding in someone can be helpful. Human connection has been shown to be a major indicator of overall health. Attempting to form a relationship with a robot may be better than nothing, but it will lack certain aspects of a bond with another human.
Scientific research has suggested that human connection is absolutely critical for addiction recovery. Some have even gone as far as saying that connection is the opposite of addiction. Addiction, studies say, is caused by a lack of healthy human connection that causes people to turn to other things to fill that gap. People that have healthy relationships are less likely to form addictions and addicts who form healthy, supportive relationships are more likely to recover.
That’s why most recovery programs involve building a community of supportive people who understand addiction and want to help each other get better. It’s the reason that a large percentage of 12 Keys Rehab’s staff are recovered addicts as well as why clients get plenty of one-on-one counseling and are connected with a sponsor.
While talking with an AI computer program might offer some feeling of connection, a machine cannot experience addiction and does not have the need for human connection. This attribute is something that only another person can offer. Mental health specialists are especially equipped to help people form the connections they need, manage their addictions and improve their mental health.
As things stand today, artificial intelligence may be helpful for mental health and addiction issues, but is not a substitute for in-person, professional treatment. Chatbots offer some benefits, such as ease of access, but also lack some of the beneficial aspects of face-to-face therapy and treatment.
It will be interesting to see where this technology goes moving forward. AI, especially in the medical field, is relatively new, and we have yet to see anything near its full potential. Reaching this potential is a significant challenge and doing so safely and responsibly creates issues as well. AI has demonstrated that it has many potential uses in the medical field, including for mental health. With time, we’ll learn more about how effective it actually is.
At 12 Keys Rehab, we focus on building relationships through one-on-one counseling and creating a supportive environment. If you’d like to know more about what we offer, contact us today.