On the weekend I had put on a load of washing, was stacking the dishwasher and dodging the robot vacuum cleaner when it occurred to me it’s a great time to be alive! In my lifetime there have been many wonderful labour saving machines invented that make housework a breeze. As a child my sisters and I had many ‘discussions’ about whose turn it was to do the dishes after dinner. And I still remember my first washing machine, a twin tub. Now I leave most of the housework to my machines. It’s a great time to be alive!
Most of today’s appliances in the home use some form of smart technology, which the marketers refer to as ‘artificial intelligence’. But they are really just using machine learning algorithms that are applied for the very specific task they’re designed for. It will be many years yet before there’s a robot housemaid or the household appliances can talk to each other.
What is Artificial Intelligence?
I’ve heard quite a few presentations about artificial intelligence and machine learning. The best have been by speakers from the Australian Institute of Machine Learning, an initiative of The University of Adelaide.
Artificial intelligence is the marketing term used for computers and machines that mimic human behaviour. True artificial intelligence will only be possible when we have technology that can think like a human, but that’s not even close to happening yet. To quote Dr Paul Danby from the University of Adelaide, “Machine learning is coded in the powerful programming language Python. ‘Artificial intelligence’ is coded in PowerPoint!”
Machine learning is the first generation of artificial intelligence, where patterns are matched by algorithms that look through thousands of examples. Computers learn how to do a task by showing it examples from a training database.
Deep learning is a subset of machine learning and is the latest development of artificial intelligence, also referred to as neural networks. These neural networks are trained using massive amounts of sample data, to develop intelligent behaviour. Once trained, deep learning algorithms can infer decisions based on imperfect and incomplete data, making what seems like us to be intuitive leaps.
Examples of artificial intelligence
There are many examples of machine learning being used in consumer electronics, household appliances and in business. There are devices that interact with us, including Siri, Alexa and Google Translate. Complex patterns in data can be identified in finance, business administration and science applications. Devices such as driverless vehicles and robot vacuum cleaners can ‘see’ their environment, learn and react on the go.
Natural language processing is an exciting development in machine learning that’s unlocking the value of unstructured data. We can use optical character recognition to extract information from scanned documents, to extract and categorise useful information quickly and efficiently. Emails with meeting minutes attached are another source of unstructured data, the minutes will have a record of who attended the meeting, when and where it was held. Information can be extracted and categorised from text documents stored in a file server. All of this extracted data can be analysed for patterns that wouldn’t be obvious to humans and can provide some intelligent insights into how well a business is performing.
Artificial intelligence that’s available now
You may not be aware of it, but many software services for business are already making use of machine learning. Accounting packages like MYOB use machine learning to automatically apply spending categories to transactions. Google and Facebook advertising use machine learning to target advertising to your customers.
If you’re keen to develop your own deep machine learning neural networks, IBM, Microsoft and Amazon all provide machine learning as a service. These have simple interfaces to help you set up the neural networks. The majority of the work is in finding and providing enough data to train the neural networks. Data is the most valuable asset to any business, evidenced by the Coles-Myer group selling off its retail business but retaining ownership of the flybuys database.
What are the possibilities for your business?
Does your business have data that is underutilised and could have its value unlocked by the power of machine learning?
If you’d like to talk further about the possibilities for your business, get in contact with me today, I’m always happy to meet and have a chat over a coffee.
Australian Institute of Machine Learning research, available at https://www.adelaide.edu.au/aiml/our-research
IBM Deep Learning – try Watson Studio for free, available at https://www.ibm.com/au-en/cloud/deep-learning
Microsoft Machine Learning Blog, available at https://blogs.technet.microsoft.com/machinelearning/tag/deep-neural-networks/
Deep Learning on Amazon Web Services, available at https://aws.amazon.com/deep-learning/