Artificial intelligence in the cloud

What we mean by artificial intelligence

Let’s start with a base definition of what we mean by AI in this context. For the purposes of this article, artificial intelligence is referring to machine learning technologies, conversational UI elements and speech recognition. For the most part, we’ll focus here on machine learning because this is where a lot of the real power of AI lives and is key to the future of artificial intelligence.

There’s a powerful saying: “AI is whatever hasn’t been done yet.” While we can’t take credit for that phrase (it belongs to Larry Tesler), it’s a great starting point for discussion. There was a point in the past where we considered some of today’s emerging technologies — things like OCR, speech recognition, driverless cars — as needing real, full-fledged artificial intelligence in order to be achieved. While AI has been part of achieving those things, we ultimately are reaching some of those milestones without a true AI computer or program that could pass the Turing Test.

AI cloud providers breakdown

So, with all that in mind, let’s discuss the three big providers of cloud technologies for AI systems and what they are bringing to the table. First, Amazon. Amazon was the leader in what we consider cloud tech today. When it began making storage and virtual machines available through APIs for customers, just as it had for itself, this was the beginning of the cloud revolution.

Fast forward to today — its entry-level AI product is called SageMaker. SageMaker has a number of AI algorithms available right out of the box. These algorithms mean you don’t need to code yourself to get started and offer a pretty simple way to get your feet wet. SageMaker also provides a lot of documentation and supporting content around the machine learning space, helping you understand what the algorithms do and how its tool works. Beyond SageMaker, it also offers Amazon ML, which is its large, full-scale and highly customizable machine learning technology.

Google’s simple AI product is called Google Cloud AutoML. It is currently in alpha, so in order to use it you need to get into Google’s alpha program. The focus of this product is largely on image processing, which shouldn’t come as a surprise when you think about Google’s involvement with image processing and driverless cars. The next level up product is Google Cloud Machine Learning, which is really its full-scale, highly customizable offering. That product is fully available and is not in alpha.

Finally, Microsoft has Azure Machine Learning Studio, which is its experimentation and entry-point product. The next tier up is Azure Machine Learning Services — a full, highly customizable production tech product. The company’s Azure Machine Learning Studio product boasts a very full set of algorithms and a drag-and-drop interface for working with those algorithms and their data sets. It’s great for someone who may know the basics already and wants more power, but isn’t quite ready for a full-blown AI system.

So, how do you go about picking the right technology for your needs? It really depends on what you’re looking for. With Amazon, SageMaker is powerful and helps developers and others get started. It has some great documentation for those who are interested and have a technical background but may not have any experience in machine learning in particular. There’s lots of information there to help you get started, not just with its tool, but just providing you general knowledge about the machine learning space.

With Microsoft, Azure Machine Learning Studio has a great user interface. It’s easy and powerful. It has built-in tutorials which help you with the use of the tool. Another powerful element of Azure Machine Learning Studio is its significant breadth of algorithms it provides you out of the box.

Google AutoML, as mentioned above, is currently in alpha, so it’s not yet a fully released product. Its specialty, perhaps not surprisingly given Google’s background with image processing and self-driving cars and so forth, is heavy on image recognition and classification.

To decide between these, we recommend you start by looking at your current technology investment. If you have a deep investment with Amazon, Microsoft or Google, you will probably be best served by choosing the AI platform that is complementary with your distinct technology.

If you choose to differ from your current technology stack, you’ll have to consider some challenges that may come out of that decision — things like network transmission, as well as the fact that the skills of your organization, as they relate to managing your existing cloud infrastructure, will translate better if you choose the platform that matches what you already have. The reality is, however, all these companies’ offerings are powerful and robust and are built on the backbones of some of the most respected tech companies in the world.

The future of AI cloud adoption

So, what’s next for AI in the cloud? We’re going to see more of this kind of tool adoption both for business and developers. As that adoption increases and more people learn about AI and what it can do for them, the skills that people develop will in turn drive even more features and capabilities as we get into the rhythm of a virtuous cycle. All this means is that there will be a high value placed on the skills needed to work with these tools. Additionally, a strong trend in the industry is that you can also expect to see more adoption of not just machine learning, but AI for things like chatbots and conversational UI.

Artificial intelligence and machine learning are certainly powerful forces in the wave of technological changes impacting industries of all types. But what we’re seeing now is just the tip of the iceberg. By choosing an AI cloud technology that will give you new ways to gather and interpret data, your organization can revolutionize the way it does business and fundamentally change the way your employees work. As Industry 4.0, the internet of things, smart homes and cloud computing technologies expand the need for artificial intelligence, we will continue to see the things that were the science-fiction stories of the past become the realities of the present.

All IoT Agenda network contributors are responsible for the content and accuracy of their posts. Opinions are of the writers and do not necessarily convey the thoughts of IoT Agenda.