What is AI, and why does it matter?
Over the recent months, you have probably seen a lot about AI or Artificial Intelligence. Whilst the images of the Pope playing basketball are entertaining, there is still a lot to learn.
Usually, AI goes hand in hand with evil villains in science fiction films and literature. Whilst Terminator and I, Robot made convincing villains; the reality is not quite as exciting. At the end of the day, AI is made of algorithms, and algorithms are just, well, algorithms. In simple technical terms, it is a complex set of codes for a programme to follow.
So what is AI really?
Artificial Intelligence (AI) is more of an umbrella term used to describe a program that is capable of some form of cognitive function (e.g. perceiving, reasoning, learning, problem-solving and interacting with an environment). Although it does not have an official technical definition.
However, there are a couple of defined categories (types) of AI defined by their abilities. Think of it like Russian dolls. All deep learning machines are artificial intelligence but not all artificial intelligence is capable of deep learning.
Machine learning (ML)
Machine learning is made up of algorithms that are 'trained' on data. These algorithms 'learn' the patterns in the data it is presented with and make predictions and recommendations. It can do this by processing data from the 'experience' it 'learns' whilst 'training'. Ultimately making better decisions over time as it gains more 'experience' (from new data and interacting with it).
This is in contrast with traditional (non-AI) machines that rely on receiving explicit programming instructions.
It might be a surprise to hear that machine learning is nothing new. The first record of machine learning was in 1943. In 1952 the first computer learning programme was written by Arthur Samuel. It was designed to play a game of checkers.
By 1997 things had advanced even further. IBM's Deep Blue shocked the world by beating World Chess champion Garry Kasparov.
Machine learning is defined as an algorithm adapting in response to new data and experiences to improve its efficiency over time. You will already be interacting with ML programmes every day.
e.g. Google Maps uses location data from your phone to inspect shifting traffic. The app can also organise user-reported traffic, accidents and construction works. By assessing the relevant data, Google Maps can reduce your commuting time and recommend alternatives.
For us at Livewire Marketing, we use Machine Learning as part of our SEO strategy for our clients. We use ML every day to help us spot opportunities and gaps in the market. This could be finding relevant content topics or spotting competitive gaps. We use Artificial Intelligence to be in the same spaces that our prospective clients are.
Deep learning (DL)
Similar to ML, Deep Learning programmes go further. They can process a wider range of data such as images, text and music. This type of AI has grown in popularity (or notoriety) in recent months. Deep learning algorithms often require less human intervention and can produce some credible results. The key word here is credible.
There are three types of DL, Feed-forward neural networks, Convolutional neural networks (CNNs) and Recurrent neural networks (RNNs). They all use artificial neural networks to collect data and process data.
This is done through multiple iterations that 'learn' the complex features of the data. From there, the neural network determines and learns whether its definition and understanding is correct. Then it uses what it has learnt to make determinations about new data.
e.g. A deep learning model can 'learn' what an object looks like and then recognize that object in a new image.
An example of everyday deep learning is something like a virtual personal assistant (Alexa, Siri etc.) Virtual assistants collect data from the user's patterns and questions. This data is then stored and used to send commands to other resources, such as phone apps. Over time your assistant will collect and refine the information on the basis of your previous interactions. This is all then tailored to suit your preferences.
Generative AI
This is more of a sub-category of deep learning. This type of AI model generates content in response to a prompt.
The famous AI tools, ChatGPT and DALL-E are both Generative AI models. These AI tools have the potential to change many industries. Although the full scope of its potential impact is still unknown, there are many concerns.
Myth busting
Stories of AI have become increasingly popular online. Unfortunately, maybe inevitably, this has resulted in the rise of wrong or purposeful misinformation.
AI will take our jobs
Whilst technological advances have always come with fear. Specifically, the possibility of machines becoming more intelligent and efficient than humans, leading to mass unemployment. Are we all going to be made redundant or worse, subject to 'them'?
Whilst AI technology can do many things, let’s just relax for a moment; it does not spell out the doom of humanity. Technological advances have historically been followed by new opportunities and industries. When computers first came about, they were thought to be the end of the world. However, you are currently sitting at a machine (or holding a device) reading about a subject that didn't exist 20 years ago.
The likelihood is that AI will be used alongside daily work tasks. Simple data jobs such as data input and sorting rolls may become redundant. However, a majority of jobs require a lot of cognitive tasks. Something humans are currently much better at than AI systems.
AI will control the world
When science fiction becomes reality, the line between stories and facts can become confused. AI programmes are tools that have limited intelligence built for their specific programme purpose. They cannot hack or break into security systems, and there is also no way to 'train' narrow AI to lead.
AI is approaching human intelligence
Although AI programmes are getting better at completing more complex tasks, such as generating musical melodies or playing games, they remain only capable of narrow intelligence.
These programmes have been built to recognise patterns based on human guidance through written algorithms. An AI programme does not choose to do a task, e.g. music; it is directed to do it.
Using the music example, an AI system is specifically programmed to learn patterns in melodies. Then it can be asked to use that knowledge to generate a similar pattern from the information it has. That same system cannot then go on to create the words to the melody or the artwork for the album cover.
How to tell the difference between AI and non-AI?
Artificial intelligence has become a bit of a buzzword within the marketing, media and entertainment sectors. When seeing anything on the internet, it can be hard to know what you are looking at.
Below, find our handy tips to help you tell the difference between AI and non-AI.
- The thing that sets AI machines apart is that they are 'trained' with data and algorithms. However, smart machines (not AI) rely on present algorithms. The 'training' allows the AI programme to make adjustments according to the knowledge it has gained. Artificial intelligence can come up with a unique solution. However, a smart machine follows a fixed programme that will give you the same result every time.
- Programmes using AI will have information about how their programme is using AI. Check if the programme you are interested in has clear information about how they are using AI technology.
- AI-based programmes capture and process data in real time. However, non-AI programmes use tags or triggers to advance to the next step of their programme. Much like an automatic door, it appears to be AI, but it actually using a sensor.
The bottom line, Non-AI is when the system keeps on following the defined set of rules to reach the solution. AI is when the system learns from its past, overcomes its mistakes and gives more optimal solutions to the problems. - Harsh Siriah, Mobile App Developer, Published by LinkedIn
Closing thoughts on AI
The hype surrounding technology is often polarised. It’s presented as a miracle salvation or the disastrous downfall of humanity. Whilst many companies are opting to add AI machines to optimise their processes, there is still a lot of room for mistakes.
Understanding the difference between AI and non-AI is important when introducing new processes. It is also essential to be able to distinguish facts from hype.
The identity of Artificial intelligence has been cemented as a singular dominant ghost. In reality, AI is a type of complex programme specially designed to carry out a specific task.
We can not unlearn or re-wind progress, AI programmes are already popular and in regular use. To what level their presence will change industries is yet to be seen.
Will we see mass unemployment and a fight to save humanity? Unlikely. More plausible is that we will grow industries and develop using AI as a step on the ladder of evolution.
Are you struggling with SEO and optimising your website? Contact a member of our team. We have a dedicated team of Digital marketing experts, from social media marketing to website development; we've got you covered.
