Artificial Intelligence: An introduction

Artificial Intelligence (AI) is a phrase you'll hear a lot at the moment, but not because of the 2001 Steven Spielberg movie but the amazing technology revolution that's swiftly becoming a reality.

This brief guide will give you the low-down on what AI is and how its currently being used


What is AI?

AI is a branch of Computer Science which aims to build machines and software that is capable of performing tasks that usually require human intelligence.

An AI agent is a piece of software than runs AI to perform a task. Multiple AI agents can be setup together to each handle small tasks in groups making them more usable and allowing them to perform much more complex jobs; similar to having a hoard of minions doing lots of little jobs to keep a company as a whole running.

In recent years AI as a whole has gone from a futuristic research topic into fully formed reality that is now being used in our everyday lives, it helps in internet searches, image recognition tasks and is also used to make predictions and suggestions.

Why is it suddenly everywhere?

It may seems at the moment that AI has exploded onto the scene suddenly, however the research into AI has been going on for decades. The reason its been talked about so much now, is the dramatic reduction in computing costs; which has made AI more viable.

AI is very computationally expensive, requiring a lot of computing resources in order to learn from large sets of information and make millions of decisions based on that information. The price has usually kept its use focused on extremely specific tasks; however because computing costs are getting lower and lower it has given the opportunity for more generalised AI such as Chat GPT or Google Bard which can perform a broad range of tasks.

We expect more and more of these more generalised AI agents to arise as time goes on, as well a more and more AI being implemented behind the scenes on services you already use.

How does it work

How AI works is a huge and complex topic, however in its most basic form an AI is created by teaching a computing algorithm about a subject using hundreds to millions of examples to create a model of that data. AI can then use that model to make predictions for whatever task is has been trained for.

Obviously this explanation is extremely brief however without going into detail of Support Vector Machines, K-means clustering and Random Forests; which even then is only a few of the algorithms used in AI learning, the above analogy is useful for a very basic understanding.

What's new in AI?

The field of AI is growing continually at pace. Chat GPT version 1 came out to the public in November of 2022 and they are already on their 4th released version with each version having increasing its AI capabilities. However the following is a list of the latest trends in AI:

Image of the author created by Gencraft AI
  • Generative AI: Generative AI agents allows you to create things from a simple prompt. Ask a generative AI focused on image generation for a picture of a "monster sitting at a laptop" and you'll be given a completely original creation by the AI. See the provided example. Generative AIs can be used for generating images, video and online content like social media posts.
  • Explainable AI: AI is complex, many times we have no idea how the AI has come up with the results is has. Enter Explainable AI! This is like a teacher who explains why they gave you an F on a test. It helps us understand how AI works and makes decisions, so we can trust it and use it safely and learn how to make it better.
  • Safe AI: You've all read the stories or heard on the news how terrifying AI can become; however there is already research into this topic and how we can keep AI a tool and not an overlord. Safe AI is like a lifeguard who watches over our AI systems to make sure they don't cause problems. It helps us prevent AI from doing bad things, like making mistakes that hurt people or the environment.
  • Federated learning: AIs need to be taught, and it can be difficult determining what to teach an AI agent, finding reliable data and ensuring that data is legally allowed to be used and not personal. Federated learning is like a secret club where people share AI ideas without telling anyone their secrets. It helps us protect our privacy and keep our data safe, even when we're using AI together.
  • Neuromorphic computing: The human brain is amazing and we classify ourselves as intelligent already, so why not build computers like the brain? Neuromorphic computing takes that idea and models its hardware and software systems like the human brain and nervous system. It produces designs that are faster and more efficient computational systems which is perfect for running AI agents.


What can I use it for?

Apart from the services detailed above, we at Digital Trading have already been integrating AI into projects, for making suggestions on input data, analysing and reporting and creating intelligent chat bots. If you have an idea on how AI could improve your business or just want to know more then get in contact with us and we'd be happy to talk to you.

Previous Article Development Lifecycles: Understanding the Key Differences Between Waterfall and Agile Development