Artificial intelligence (AI) and machine learning (ML) work together to make smart decisions and solve problems.
AI services are the most accessible and require the least expertise, making them suitable for users who want quick, out-of-the-box AI capabilities.
ML services offer a balance between ease of use and customisation, catering to users with specific data needs.
Machine learning frameworks and infrastructure services provide the highest level of control and are made for experts who need full flexibility and customisation in their machine learning models.
What are artificial intelligence (AI) and machine learning (ML)?
Artificial intelligence (AI) is the magic that makes computers think and learn like humans.
It lets machines understand, reason, and make decisions, and have human intelligence.
Machine learning (ML) is a subset of AI. Think of AI as a wide umbrella of different technology, and ML is just one of them.
ML is like a teacher for computers. Instead of explicitly programming every step, ML systems learn from data.
ML gets better at tasks with practice, like recognising faces or predicting a company's sales numbers.
AI and ML often go hand in hand - it's rare to see a use of AI that doesn't involve ML at all!
Think of AI as the brain and ML as a way for the brain to learn and get smarter.
Just like how you learn from your experiences, ML lets the computer learn from data and improve over time. Think of ML as a coach for AI to practise and be even more useful!
AI/ML services in AWS are expert assistants that save you from building everything from scratch. They provide pre-built tools and algorithms that can analyse data, recognise patterns, and make predictions.
Who uses AI/ML services and why?
People of all kinds of backgrounds and industries use AI/ML services in AWS. These services help businesses make smarter decisions, automate tasks, and gain insights from data. Let's take a look at a few examples:
Netflix/Spotify (AI for content recommendations): When you enjoy your favourite shows on platforms like Netflix and listen to music on Spotify, you benefit from AI. These services use AI algorithms to analyse your viewing and listening history and recommend new shows, movies, and songs that you're likely to enjoy. This personalisation keeps you engaged and entertained, which is a big win to the companies too.
Apple Photos/Facebook (ML for photo tagging): When you upload photos on Facebook or save them in your iPhone, these platform uses ML to suggest tags for people in the photos. ML algorithms analyse facial features and previous tags to make these suggestions, making photo sharing and organising more convenient.
Apple/Amazon (AI in customer service): You may have spoken with Siri or Alexa (i.e. Apple and Amazon's virtual assistants) to check the weather, call someone, set reminders, or play music. AI-powered technology is what makes these virtual assistants understand what we say or text, and get information back to us so quickly!
There are three levels of AI/ML services...
We've learnt that different kinds of AWS services provide different levels of support. For example, you'll be managing much more if you're spinning up EC2 instances, while you'd only need to think about your code if you're making Lambda functions.
AI/ML services are no different. Here are the three 'levels' of services, which means they give you a different level of control and customisation:
AI services provide pre-built, ready-to-use AI models and tools for specific tasks, such as language translation, text-to-speech, or image recognition. No need for in-depth AI knowledge; just use them as-is! While AI services are convenient, they offer limited customisation. Users typically can't modify the underlying AI models.
ML services provide tools and algorithms to build custom ML models. Users have more control over the training process and can fine-tune the models to their needs. This means they give you a balance between ease of use and customisation, making them great for tasks where you have specific data and goals (e.g. predicting customer behavior).
Machine learning frameworks and infrastructure services: These provide the tools and resources for building your own AI/ML models from the ground up. If you're an AI/ML pro (cue data scientists and machine learning professionals!), these are your playground. You can fine-tune models and have complete control over the process.
Are there any AI/ML services in AWS?
Absolutely! AWS offers a wide range of AI and ML services to suit various needs. Here's a sneak peek at some of the popular ones:
