We hebben jarenlange ervaring met transities van organisaties naar Cloud-oplossingen. Wil je hier meer over weten neem dan contact met ons op.
We hebben jarenlange ervaring met IT security voor diverse bedrijven. Wil je hier meer over weten neem dan contact met ons op.
Most companies are driven by digitalization pressure rather than a desire to take control of their own Digital Development.
Managed Services behoort tot de kern van onze bedrijfsvoering en dat maakt ons tot de aangewezen partner voor organisaties. Wil je hier meer over weten neem dan contact met ons op.
We can better exploit new opportunities when we understand what new technologies involve and how they interact. Today’s topic is what artificial intelligence and machine learning are, and an insight into how they can relate to IoT and Big Data.
At Basefarm we work daily with these terms. Something we notice is that many equate artificial intelligence and machine learning. Another thing is that there are several different descriptions of each, and those definitions even change over time. So, if you choose to dive into this, you are best off to study fresh sources.
Here is an attempt to clear things up.
Fortunately, the terms themselves help us out. We humans have intelligence. We take in sensory impressions – data – from different sensors like skin, eyes, ears, mouth and nose. From birth and onwards, our brain has been programmed into an intelligence we use for problem solving with subsequent actions.
Artificial intelligence (AI) is an analogy to our own intelligence. It is one or many programmed computers that process collected data, make decisions and perform actions.
Just as for humans, intelligence can vary. Machines with narrow artificial intelligence can complete specific processes and nothing else, such as recognizing a face in an image. Machinery with general artificial intelligence can perform more like the human brain, with problem solving based on various inputs and a broad span of possible actions.
Some define the limit of general artificial intelligence as the limit of the human brain capacity. The next step would be an artificial superintelligence which can perform even more than the human brain. The idea of an artificial superintelligence might appear promising, frightening or perhaps even somewhat comic, like in the sad case of Marvin in Hitchhiker’s Guide to the Galaxy, a robot that is struggling with depression.
But, let’s return to our real world and the next term on the list: machine learning.
Again, the term itself helps us out. It is about machines that learn. And, the term is closely intertwined with AI. As people learn from new inputs, so can machines – which enables them to make better decisions.
A simple programmed AI will repeat its processes over and over again. An AI with machine learning capabilities provides a greater basis for problem solving and evolves during time. This way, machine learning can be seen as a component of artificial intelligence.
But, the question remains; what can AI be used for? HAL 9000 in the Stanley Kubrick classic 2001: A Space Odyssey learned to sing during his training, was able to control a spacecraft, read lips through a video camera and retrieve data from other sensors, keep secrets, and send messages through voice, monitors and other means of communication.
Likewise, AI can gather data from sensors. In our world, many of them come from the Internet of Things (IoT). AI can provide information and commands in return to the same or other IoT networks. This way, AI and IoT are closely interwoven as brain and body are, a combination which many believe may become the basis for an even deeper learning than machine learning.
Obviously, as AI and machine learning represent different processes, they need different software and hardware. Machine learning demands infrastructure that can gather and process big data oceans, which for one thing takes a lot of processing and storage. Artificial intelligence has a large need of processing capability, including internal memory access.
The workloads usually vary over time. So, in order to build a scalable AI solution you need to build it on flexible infrastructure that leaves room for growth.
With our vast experience in building and running mission critical applications, we at Basefarm are your ideal partner to plan, set up and securely run highly scalable infrastructure. Also, the physical distance between IoT and AI should not be too long as the communication should be high speed. With our own data center proximity and partnership with major public cloud providers, Basefarm is in the best position to handle this kind of on the edge computing so you can get the most out of the modern, ever-evolving AI and machine learning technology.
How to become a sovereign data enterprise – step by step.
Download the white paper here!
Find undiscovered secrets in your data and on the web: intelligent algorithms provide you with unique knowledge about your customers and your business.