Scientists develop a new digital assistant named after the famous detective Sherlock Holmes.
There is a new digital assistant beats Siri and Cortana. Scientists at the Cardiff University showed off the next generation voice assistant running on smartphones and the Apple Watch.
Don't Miss: Nintendo Switch: Everything You Need To Know
SHERLOCK stands for Simple Human Experiment Regarding Locally Observed Collective Knowledge. The system is using a novel "controlled natural language" which ensures that the software and its users understand one another, enabling Sherlock to act on the information it's given in a robust way.
Unlike most digital assistants, users can tell SHERLOCK things and it can use the information it's told. It combines it with things that other users tell it to build a "local knowledge base" allowing it to work out new information and take actions.
Researchers say SHERLOCK can therefore be used to help people communicate more effectively with computerized systems than ever before. For example, in a smart home you could tell it, "I'm feeling cold", and it could adjust the heating or suggest closing open windows.
Similarly, it can also be used to collect and process 'crowdsourced' information. At a festival it could help people work out where the busiest and quietest food areas are, or in an emergency it could help medical teams – or even aerial 'drones' – to find people needing aid, based on eyewitness reports.
Over the course of the next year the research team will carry out a series of experiments in collaboration with US psychologists to build evidence that the approach really does help people and machines to work better together in challenging situations.
If successful, the team hopes the technology will progress into the user's hands.
Professor Preece demonstrated the Sherlock at the BBC Make It Digital event in Cardiff.
Don't Miss: The Best HDR TVs
We assume Apple and Microsoft will take note of Sherlock and will see how this aligns with their future development plans for Sire and Cortana respectively.