Read about artificial intelligence, machine learning and cognitive computing.
Are artificial intelligence, machine learning and cognitive computing the same?
Artificial intelligence (AI) embraces machine learning and cognitive computing, but they are not the same.
- AI is one specific branch of computer science that attempts to create machines that are capable of acting intelligently.
- Machine learning, can be defined as the science of creating computers that act of their own volition, without explicit programming.
- Cognitive computing is when a computer is able to simulate human thought processes. It involves systems and algorithms being learnt that can then recognise patterns, mine data and process natural language to mimic the brain.
They are similar as AI researchers are necessary for building smart machines, but experts in machine learning are necessary to make those machines genuinely intelligent.
The foundations of artificial intelligence
The term artificial intelligence dates back to the 1940s when Alan Turing developed the codebreaking machine and subsequently the now famous Turing Test, which was based on the concept that a machine was capable of thinking.
The codebreaking machine which helped to end World War II, used machine learning to identify the codes needed. Today, the major players in tech, like Nvidia and Google, are working on developing machine learning and pushing hard for computers to learn to think the same way humans do so that they can progress into what is being hailed as the next technological revolution: machines that can ‘think’ like people, that’s cognitive computing.
In the last decade, machine learning has led to innovations such as self-driving cars, far superior web searches, practical speech recognition and a massively improved understanding of the human genome.
How exactly does AI work?
At its most basic, AI and machine learning are exemplified in the handy way in which Google picks up on the fact you made a typo and changes your search parameters to what it thinks you meant instead. A little note pops up saying ‘Did you mean…’ or ‘Showing results for…’, giving you the option to stick to your original (potentially wrong) search or see the results for what Google thinks you really meant. Nine times out of 10, Google is right and you made a mistake. This machine learning algorithm is a small thing, but it saves people a lot of time trawling through search results that aren’t relevant because they failed to notice they made a spelling mistake. Google’s search results are predominantly delivered through Rankbrain which is a form of artificial intelligence.
Artificial Intelligence in practice
Here are two examples of artificial intelligence in practice:
- Expedia – the airline provider’s chief data officer, Matthew Fryer, described AI as ‘the core of why we exist: to find the right holiday, flight, and experience for you’.
- Moo.com – the specialist personalised printing company use AI to improve their customer services by harnessing text analytics to read and analyse text from customers. This allows the company to prioritise responses to messages. In the future, this will enable the company to automate some responses and highlight those needing urgent personalised responses due to sentiment analysis – an indicator of positive, negative or neutral comments from customers. Moo also use basic machine learning for providing better on page search results. They focus on specific keywords and if many people search for certain content on that page, it becomes a more popular result and more likely to emerge when the next person comes on and searches for something similar.
Machine learning in practice
Mainly focused on travel, two examples of machine learning include:
- The private taxi company Uber use apps as the core of their business. Need a cab? Use the app! They use machine learning to better predict their customers’ travelling habits and improve their maps to ensure the most efficient routes possible.
- Your satnav also adapts based on your favourite journey, it learns which routes you take and ignore and presents future results based on your driving habits.
Cognitive computing in practice
Cognitive computing in healthcare automates processes, such as computers collating all possible knowledge surrounding a particular condition. This may include the best practices and treatments advised, what the patient’s history is and how this could impact the condition and what academic evidence is available. This enables doctors to provide evidence-based treatment plans based on the individual and the latest research available.
Does your business use artificial intelligence? Do share how and we can include you as a case study.