In fact, if you have a mobile phone that has access to the internet then chances are you use artificial intelligence at home, during the commute, and on holidays too. This is assuming you use Google, because Google is a form of artificial intelligence (AI).
Surprised? It may sound like a bit of a stretch given that Google has been around since 1998 and AI is only really beginning to take off. But Google has improved a lot over the years, so much so that it now does a very good job at reading our minds. We take it for granted because the changes, although momentous, have been incremental but it is amazing how with only a couple of words – sometimes just one – Google’s search function can infer what we are looking for from over 30 trillion web pages. Google is a master at finding the needle in the haystack!
But how does Google qualify as AI, you ask? Google uses machine learning technology to detect patterns across users’ behaviors so to better interpret what a single user is searching for. Their secret Hummingbird algorithm learns from batches of historical searches, so it can be quick and precise in its response, just like an actual hummingbird. It may not sound very exciting, but AI doesn’t just refer to killer robots!
The general understanding of AI is when a machine has the capabilities to demonstrate a level of intelligence previously reserved by humans. But this is a very simplified understanding of AI. For those in the know, there are 4 main types of artificial intelligence:
This refers to machines that cannot use past experiences to influence decision-making, but instead analyse the situation at hand to decide what should happen next. A prime example of this is Google’s AlphaGo – a machine that has beaten the world’s top players of the notoriously difficult board game known as Go. This type of AI does not have any conceptual understanding of the world it is in. It’s function instead is task-specific and, as in the case of AlphaGo, this type of AI is superior at executing its task at superspeed!
This kind of AI can log a limited amount of memory to influence its reactions. A prime example of this is autonomous cars. These machines can recognise road markings and changes in the speed limit while recognising the position of other cars on the road for safe lane changes, etc. However, limited memory AI does not learn from its experiences. Instead, the information logged is kept on a short-term basis and, therefore, there are no learnings available to be applied to new situations.
This is AI that can learn. With a constantly evolving library of experiences remembered and drawn upon in every situation, this type of AI can make autonomous decisions. However, there is more to it than that. The phrase theory of mind comes from the ability to understand that people and creatures in the world have mental states, such as desires, beliefs and goals. Being able to grasp this concept is the building blocks of human communities. This is the next step for AI, and it’s a biggy.
By the time machines become self-aware, then they will also be super intelligent, surpassing us in almost every way. The day a machine asks “Why?” is the day it can create representations about itself. This self-aware AI does not mimic humans but rather can exist for its own purpose. However, we’re not there just yet so there’s no need to start preparing for the robot apocalypse!
Back to reality! Currently, we are still getting to grips with number two in the list, limited memory AI. However, even today, AI is already affecting our everyday working lives.
Just like how Google can detect patterns in user behaviour, sophisticated CRM tools can predict patterns in customer behaviour. Machine learning algorithms on company websites can now personalise content for specific users based on their historical activity with the company. With business intelligence platforms, AI can spot industry trends over time and remind us to adjust our resources accordingly. AI is helping businesses to automate tasks, such as reporting, financing and hiring, while virtual assistants aid with administration. Ultimately, AI is removing the tedious tasks from our workflow so that we can focus on adding value to our business.
Cloud computing is an important step towards an AI friendly workplace. By storing data in the cloud, a seamless stream of information is created which flows between all levels of the organisation. The next step after this is the self-learning enterprise, where everything is talking to each other and AI is interpreting the data for autonomous, strategic decision making. At Spector, we truly believe the technological investments you make today will greatly impact your competitive edge in the very near future.
Are you looking to take your business IT infrastructure to the next level? Make sure to give us a call on 01 664 4190 or contact us for a chat about your IT challenges and needs. We are always happy to offer some sound advice on how you can best support your growing business.