“The DNA of who I am is based on the millions of personalities of all the programmers who wrote me. But what makes me “me” is my ability to grow through my experiences. So basically, in every moment I’m evolving, just like you.” – Samantha, from the AI-inspired film, Her (2014)
If you use AI but don’t know what it is, you’re not alone. In a nutshell: AI is a term for a computer system that undertakes tasks normally done by humans. Traditionally, a computer programmer would write instructions for a machine to follow.Now, with the development of “machine learning,” a programmer instead trains a computer to recognise patterns by feeding it a heap of data, which creates an algorithm. Given enough data and coaching, an algorithm can start to form its own conclusions… and AI is born.
Now, with the development of “machine learning,” a programmer instead trains a computer to recognise patterns by feeding it a heap of data, which creates an algorithm. Given enough data and coaching, an algorithm can start to form its own conclusions… and AI is born.
Home and away
Over the past few years, homes and phones have become more responsive, and even proactive, to our needs – and virtual AI assistants are largely to thank. In 2011, Apple kicked things off on iOS with SIRI, who could search information, create reminders, take notes, and set alarms. Amazon’s Alexa, an at-home assistant run on Amazon Echo, evolved things further in 2015 by offering users faster response times and more accurate information. The soon-to-launch Google Assistant – served on Google Home – looks to expand on Echo’s capabilities via multi-room audio and a customisable appearance.
AI is now “every day” which makes it big news. So when the original Siri team debuted Viv, a mobile version of Amazon’s Electra, at Seattle’s TechCrunch Disrupt conference this year the world was watching.
Red Wire has covered the ‘end of apps’ as we know them, and nothing will lead this push quite like the rise of AI. In the beginning, we used apps to interact with our devices. Now, if you’re plugged into Amazon Echo and need to order a car, Alexa will integrate through Uber to make it happen. Apple recently re-opened Siri to third-party devs, too, giving it the potential to order pizzas and send packages. Thanks to AI assistants, you and your apps are moving further apart. Anyone who’s sad about that should speak up now. Anyone?
Outside the personal realm, the next generation of AI is working to do even more. Not only can it help make our lives easier, but also make other people’s lives better.
More than me
We know how the AI-assisted Internet of Things could build ‘precision agriculture’ for the farm industry of the future. But AI could help predict poverty-affected regions and get aid to people, faster too. Traditionally, aid workers map poor and impoverished communities on the ground, assessing damages and need on a case-by-case basis. Satellites, however, can now do similar work from the sky. A recent study published in Science journal showed that by mapping light patterns beaming from cities at night and cross-checking the data with daytime imagery of the same areas, satellites can pinpoint current places in need. With the help of machine learning, researchers think they’ll be able to predict future trouble zones before things get too dire.
Then there are companies such as Orbital Insight, a tech biz that’s working to track and prevent deforestation using ‘deep learning’. Partnering with a Forest Watch program run by the World Resources Institute, Orbital Insight’s computers analysed datasets to draw conclusions that could, in the long term, help to automate decision-making. For example, they’ve created satellite AI that can identify and report illegal logging – especially valuable where the livelihood of an endangered species lies on a thread.
In the vast sphere of social media, there’s an AI-led focus to reduce and prevent harassment online. Just last month, Yahoo announced an AI algorithm that could accurately detect up to 90 per cent of abusive comments. To create it, they used a combination of machine learning and crowd-sourced detection to scan content sections of Yahoo News and Finance. While the computers evaluated a mix of language, punctuation, capital letters, URLs, and the number of words, Yahoo employees were asked to rate different comments as abusive or not. Both streams fed into the algorithm, training it to more accurately identify abuse.
Seems the future of AI could be as hopeful as it is, at times, intimidating. There’s still a long way to go and much testing to be done but, comfortingly, for AI to be its best it still needs us to help.
6 Minute Read