Imagine going to a doctor, describing your symptoms, and, after a series of follow-up questions, learning your diagnosis and treatment plan. And all without ever meeting a doctor. Although it may sound incredible, it is already a reality in California. Startup Akido Labs, with its ambitious ScopeAI project, is slowly beginning to transform healthcare.
ScopeAI: A “Doctor in a Box” That Saves Time
ScopeAI is a system that uses large language models such as Llama or Claude to conduct basic medical examinations and make diagnoses. In practice, this means that while you meet with a medical assistant at a healthcare facility, all questions about your health, proposed diagnosis, and treatment are determined by AI. A doctor then merely reviews and approves the results or adjusts them if necessary.
Among the first patients testing the new system are people insured through Medicaid, a program intended for low-income individuals and families. They are often the ones who face inadequate healthcare.
Accessible and Timely Care
The American population is aging, and cuts to federal healthcare funding are making the entire situation even worse. Another problem is the shortage of doctors, who are unable to provide timely care to everyone. ScopeAI aims to change exactly that.
According to experience so far, AI allows doctors to examine 4–5 times more patients than before. ScopeAI is behind this increase in productivity, reviewing symptoms, making a diagnosis, and creating a treatment plan. Doctors then only need to review the results, save time on routine work, and can immediately focus on serious illnesses or lower-income patients.
Risks and Legal Obstacles
However, some experts argue that the new “AI assistant” does not solve the problem of doctor shortages or inadequate healthcare. They are also concerned that doctors may rely too heavily on AI, which could lead to incorrect diagnoses. This also raises legal questions.
Artificial intelligence is already commonly used in healthcare today. It helps detect cancer in preventive scans, is used in cardiology and endocrinology, and even searches for information in specialist literature. In all these cases, however, it does not stand in for doctors but merely assists them. In the eyes of some experts, the ScopeAI system therefore has no legal basis and violates medical licensing laws. These laws stipulate that AI may assist doctors but not act in their place.
According to its supporters, however, ScopeAI does have a legal basis. Although the basic examination takes place without a doctor present (with only an assistant), the doctor still has the final say, and the information collected by the AI is merely a “proposal” that is then reviewed.
An AI Assistant With a 92% Success Rate
ScopeAI's primary goal is to make healthcare more affordable and accessible to as many people as possible, and so far it is succeeding. The system has also been praised by doctors who work with homeless people. With its help, they can provide medications within 24 hours, something that was previously unthinkable.
ScopeAI is designed to assist doctors, not replace them. Therefore, it always suggests 2–3 alternatives to its diagnosis, and the doctor then evaluates whether the system made the correct decision. According to the results so far and testing on historical data, Akido Labs' ScopeAI already achieves a 92% success rate.
Although the initial response has been positive, no larger-scale clinical studies have yet been conducted to show whether the results match or even surpass those of conventional examinations.
Source: MIT Technology Review
