The Beijing-based search engine is offering an AI-based medical assistant app, which it says will create a quicker path to getting information from patients to doctors.
From Siri to Alexa to Cortana, chatbots are beginning to revolutionize the way we receive information. Gartner reports that, by 2020, 85% of customer interactions will be managed without humans.
But while many bots are focused on delivering goods and services to consumers, a new bot by Chinese search giant Baidu aims to solve a different problem: Delivering assistance to those seeking medical help.
Baidu, one of China's top tech companies, is no stranger to AI. It has partnered with BMW to develop driverless technology in Beijing. And in September, it unveiled an open-source platform for machine learning.
It's new AI-powered app, AI Medical Assistant, aims to help doctors receive preliminary medical information from patients--before they are scheduled for an appointment. It's linked with Baidu Doctor, an app that offers patients a way to ask doctors questions, schedule appointments, and find healthcare information.
Once downloaded, a patient can ask a question through the app. The Baidu AI Medical Assistant, Melody, will ask follow-up questions about symptoms, to clarify the severity of the health problem. That information will be conveyed to a doctor, who will then make a decision about whether the patient needs an immediate appointment. The assistant can also suggest treatment options to the doctor.
The World Health Organization reported a shortage of 13 million healthcare professionals over the next twenty years. The problem is particularly pronounced in China, which has a rapidly aging population. And China, unlike countries like the US, doesn't have primary care physicians.
Wei Fan, director and deputy head of Baidu's AI lab in California and a lead in the development of Baidu Doctor, said that the app is not meant to replace doctors--or offer a diagnosis.
"Even though the whole system has been tested extensively by medical professionals in China, with over 200 senior medical students as well as another 20 or so experienced doctors from the very, very top hospitals, we are very careful about our diagnosis," he said. "Maybe in the future we could offer one, but for the moment, it's mainly like a medical assistant."
Others in the chatbot world see this as offering an important service.
"The Melody chatbot is an important first step taken by Baidu in revolutionizing the healthcare industry," said Alexandru Iliescu, CEO of ATi Studios and the maker of Mondly, a chatbot for learning languages. "From its infancy, Melody is clearly giving a helping hand to doctors by having an initial chat with patients and analyzing huge amounts of data in a heartbeat. Humans cannot crunch all the available data that fast, but a chatbot can. That's the beauty of AI."
Robbie Abed, director of product strategy at Y Media Labs, sees it as a "good entry point."
"As long as it doesn't misdiagnose you, I think the initial conversation of getting some more contacts is a good thing," he said. In the future, Abed said, he'd like to see vitals become integrated. "Right now it's just not a possibility with the system, but longer term, I see it as something that is viable."
But other bot developers have concerns over the technology. Amit Aghara, global head of solution management at Kore, which makes bots for the enterprise, wonders about the platform connected to the bot. In order to be successful, the bot must be able to both remember and deliver context to the conversations. "Without a platform, trying to customize and scale that solution to every health care, and what they may be looking for, is going to be extremely difficult," he said.
"The bot in itself," said Aghara, "may not be something large enterprises would just pick up and use."
- New MyKAI smart bot uses AI to enable 'lifestyle banking' (TechRepublic)
- Amazon open sources its deep learning software (ZDNet)
- How Amazon wants to bridge the data science gap by bringing machine learning to the cloud (TechRepublic)
- Machine learning face-off: Microsoft uses Band to show what its Watson rival is capable of (ZDNet)
- Google DeepMind: The smart person's guide (TechRepublic)
- How developers can take advantage of machine learning on Google Cloud Platform (TechRepublic)