In this article, FuturU’s CTO Steve Lowe reflects on the trajectory of AI, its practical applications in healthcare, and some of FuturU’s developments in this space.
AI may be the buzzword of the past year, but it’s actually been around for quite some time. My earliest recollections are of Amazon leveraging AI to predict demand based on factors like weather and time of year. For example, if the weather forecast was showing a 90% chance of rain in London, they might ship extra umbrellas to their London depot. The goal? Streamlining operations and ensuring items were strategically placed for efficient delivery. This early AI application involved vast amounts of data, pattern recognition, and predictive modelling – a precursor to the AI landscape we navigate today.
Generative AI
More recently, generative AI has come onto the scene, with chatbots and image generation being two of the most popular use cases. In principle, these are all very similar ideas based on large language models (LLMs). Chatbots, for example, use vast amounts of data to predict what the next word in a sentence should be. When that’s done repeatedly, the chatbot is able to produce sentences and long-form content. If it’s all weighted correctly by the context, the content it delivers should hopefully make sense. However, it’s not without its limitations. ‘Hallucinations’, where the content seems accurate and convincing but is in fact wrong, can be fairly common.
At FuturU, one of our biggest generative AI projects to-date is our tutor chatbot, which will provide learners with a 24/7 tutor to help them on their learning journey and answer any questions related to our training. For example, say a nurse had a specific question about phlebotomy (drawing blood). To ensure the response is factually accurate and not at risk of hallucinations, we layer information and only provide results from a fixed body of knowledge. In this case, our training course on phlebotomy. The aim is to provide reliable assistance, while also giving learners the flexibility to access training support on the go.
AI assistants in healthcare
AI can also act as a great assistant to help save time and speed up processes. Imagine you’re a GP and you’ve seen a patient with an unusual set of symptoms and you’ve hit a brick wall. Thanks to its massive repository of knowledge, a chatbot could provide you with a second opinion, refresh your memory or suggest a potential diagnosis in seconds (saving you a lot of time!). Of course, it’s important to be wary of hallucinations and fact-check any information provided, but the benefits are undeniable.
AI can also be great at repetitive tasks and assessments of statistics. For example, if you’re a radiologist with thousands of CT scans to monitor, an AI assistant could scan them in a matter of minutes, quickly detecting any anomalies. In both cases, AI is not designed to replace professionals, but more to assist or enhance their work – ultimately improving patient outcomes.
AI in education and training
Another area with huge potential is the integration of AI in education and training. At FuturU, we’re building a motion capture tool that uses a more traditional form of AI to help train people on safe moving and handling techniques. Using a camera, the AI is able to work out where your legs and joints are. From here, it can detect where your leg bends and therefore where your knee is. Once it’s able to do that for all of your joints, it can then map your skeleton and measure how far you’re bending down, which way you’re facing or whether your hips are slightly rotated. This can be a very reliable tool for assessing someone’s physical movement, and in many cases is much more accurate than human observation.
We’re also exploring how AI could be used to simulate different patients in a training context. For example, being able to simulate a patient with dementia and have a virtual chat with them, or de-escalate a situation with a dangerous patient in a prison. Practising these scenarios in the real world can be difficult, so having the chance to train in a safe environment with realistic characters and interactions could be hugely beneficial. Simulations that allow practitioners to experience and respond to complex scenarios could significantly enhance their skills, giving them more confidence in their abilities to do their job.
Ethics in AI
As FuturU’s CTO, I often find myself contemplating the ethical use of AI. While the technology opens new doors, it’s crucial to emphasise responsible and safe practices. Users must approach AI as a tool, understanding its capabilities and limitations. The potential for positive impact is vast, but a cautious approach ensures we navigate the AI frontier responsibly.
AI’s influence on our lives will undoubtedly grow. The key is to stay informed, adapt, and harness the power of AI to enhance our daily experiences. The future is exciting, and with responsible AI integration, we can pave the way for a technologically enriched world.
