The Latest Developments in Artificial Intelligence
- Published 28th Feb 2019
Last edited 28th Feb 2019
Artificial Intelligence is a field of study that seeks to create computer systems that possess human-like intelligence (such as cognition, problem-solving, speech, and speech recognition) while retaining the efficiency of a machine. The ideal AI tool will be able to think, interact, and learn like a human, but work harder and longer than any human.
While the sector is still many years away from reaching its target, there are already a number of functional AI-based systems in circulation e.g. Google’s Assistant, Amazon’s Alexa, Apple’s Siri amongst many others. Below are some recent developments from the artificial intelligence sector:
AI tool that prevents Netflix password-sharing unveiled at CES 19
Streaming services lose millions of dollars every year because of password-sharing. To curtail this, a video software company has created an AI tool that can be used to track and identify users that share their account information with others.
The system was unveiled at the CES 2019 technology trade show and it will use machine learning to analyse user data for signs of irregularity e.g. an account being used at 2 places within a short period of time.
EU travellers to be questioned with AI-based lie detector system
Hungary, Greece, and Latvia plan to use an AI lie-detector system to question people trying to gain entry into their countries. The system will assess each traveller’s documents, biometric data, and social media activity, and it will ask them some questions. Anyone that the system deems to have lied will be transferred to a human agent for further questioning.
The system, named iBorderCtrl, is said to analyse 38 micro-gestures in a bid to spot facial patterns that are indicative of lying. During testing, the system was reportedly correct 76% of the time and its creators plan to use data to further improve its accuracy.
AI robot sits in on UK parliamentary session
The first AI robot in the world with cultural awareness, Pepper, was brought before a committee of MPs to answer some questions. She gave answers according to a script she was fed beforehand and generally, gave a good account of her artificial intelligence.
Currently, Pepper serves as a receptionist in a number of offices in the UK and she is reportedly able to use facial recognition to identify visitors, send meeting alerts, arrange for refreshments, and chat with potential clients. The eventual goal, according to her handlers, is to train Pepper “to provide care and companionship” to the elderly.
AI learns common sense by playing Pictionary
Researchers from the Allen Institute of Artificial Intelligence in Seattle have designed a version of the popular game, Pictionary, to teach their artificial intelligence, AllenAI, how humans think and how to collaborate with them.
The game is called Iconary, and people play it alongside the AI so that it can gain a human-like grasp of common sense. The human player can either draw an image for the AI to guess or guess based on an image the AI draws.
These are just a few of the developments in the AI scene. Efforts to create an AI that possess human intelligence has been going on for years, and it will surely continue for many more. How close are we to building the ideal AI? That’s the billion-dollar question.