Anything you can do, AI can do better: Making infrastructure smarter

24, February, 2017


Artificial Intelligence -



Data flows and artificial intelligence (AI) are changing the globalised supply chain by sending data around the world. Using the cloud as the data transport method requires more efficient data acceleration. AI is infiltrating many places, and it’s not just about replicating human performance and taking jobs, or streamlining processes. It also helps make technology smarter.

AI can be part of IT infrastructure. David Trossell, CEO and CTO of Bridgeworks, argues that it can be a good thing, and needn’t involve employees being made redundant. In his view, artificial intelligence offers a good story as it can enable organisations to better manage their IT infrastructure, and enable organisations to improve their business intelligence, allowing them to make better decisions at a time when big data volumes are increasing at a tremendous rate. In fact, CTOVision.com claims that some experts predict that by the year 2020, the volume of digital data will reach as high as 40 trillion gigabytes.” So the task of sifting through it going to become harder and harder.

Lucas Carlson, senior vice president of strategy at Automic Software, believes it’s been an interesting summer – and for reasons most of us wouldn’t think about. He says that there are ‘5 ways artificial intelligence will change enterprise IT’ in his article for Venture Beat. He says artificial intelligence can be used for predicting software failures, detecting cyber-security issues, creating super-programmers, making sense of the Internet of Things, and managing robots in datacentres.

Rise of AI

Just a few short years ago artificial intelligence was at the beginning of the hype cycle along with fuzzy Logic and other such things, and so much was hoped for out of these technologies but they slowly faded away in to the background”, says Trossell. This pushed artificial intelligence out of favour, becoming more of an academic interest than anything else.

That was until the ground-breaking and great movie, AI, came along and shocked the world with life-like human thinking. But whilst it brought AI back to the forefront it did so with the fear that robots would displace humans as the superior race”, he explains. One could add that there have been many science fiction movies like this – such as the Terminator series of films. The films are about the battle between humans and killing machines – cyborgs – which take on a human form in order to enable them to walk unnoticed until they attack their targets with an eye on defeating humankind. Perhaps the most shocking aspect of the story is that the machines were originally built by our own race.

Trossell thinks the doom and gloom presented by these films will happen well after his lifetime.  Yet hopefully, mankind will choose to make more machine intelligence for our own good, rather than for our own destruction. Nevertheless, he thinks that AI is slowly re-emerging in a positive light, not as the prophecy of destruction of mankind but as a companion.  Many people are still concerned that this will displace jobs and lead to global unemployment with all the social upheaval and ills that accompanies it, but history teaches us otherwise.”

He explains why this is the case: At the start of the industrial revolution, James Hargreaves’ Spinning Jenny suddenly changed the lives of many cottage industry spinners, replacing many with one largely unskilled worker. Out of this there was a massive increase of employment as other industries mechanised to take opportunities with the increase in spun wool.  Yes, we did go through the age of the dark satanic mills, but industry is at the heart of our civilisation and many of us enjoy the benefits created by it with an exceptional quality of life.  What this shows us is while there is a short-term displacement in employment; this will be absorbed by new industries created later.”

Expert systems

He says that AI has the ability to create expert systems that never tire and they can augment humans.” For example, artificial intelligence is being used to spot breast cancer as it has the ability to learn – improving the cancer identification rate by up to 99.5% accuracy. AI can be employed in many other situations where it can augment experts to drive efficiency and ROI. Data is all around us – most of us are drowning in it – but it drives our modern society”, he comments. Yet this leaves the question of where to store this growing volume of data. Another question that often needs to be answered by many organisations today is about how to move it quickly and securely around the globe.

Data is at the heart of every organisation and many have a global presence as well as customers worldwide, and increased distances create a major bottleneck”, he warns. Although network speeds have increased exponentially over the past few years, this has not necessarily improved the performance of transmitting data over distance”, he explains. In his view, what it has done is exasperated the problem and delivered organisation a poor return on their investment. He says this is all caused by the ghosts that plagues networks; latency and packet loss, but these can be mitigated or reduced with the help of AI.

Mitigating latency

There are techniques that highly skilled engineers or programmers can employ to mitigate some of the effects of latency and packet loss, but the network – and more importantly wide area networks (WANs) – are a living, unstable entity to work with”, he warns. Therefore something has to be done, and preferably without much human intervention. After all, we as humans are all prone to making mistakes. Traditionally, to maintain the optimal performance of the data flowing across a WAN would require an engineer to constantly measure and tune the network, and this is where errors can be made.

With AI, the goal of maximising the performance removes the need for constant human intervention – leading to the potential benefits of reducing any human error, better network management, improve disaster recovery, reduced packet loss and increased accuracy. Artificial intelligence can dump a set of rules and learn in a similar way to how it is being deployed in the breast cancer article. It can learn from its experience the way the network behaves, and it also knows how data flows across the network”, he claims. This is enabling society to develop a fit and forget approach.

Two solutions that do this are PORTrockIT and WANrockIT. They help to mitigate the effects of latency and reduce packet loss. Beyond them Trossell believes that we have just begun to scratch the surface with the possibility of AI, and  there are many other thought and time intensive processes within the IT world that it can be  applied to, if only we had the courage to release control.” That must happen because we as humans are reaching the limit of our ability to manage the ever-increasing data volumes without some form of machine intelligence to support our endeavours. Could we hand over to AI the management of space, the location and the accessibility of data? Yes we can, because anything we can do, AI can often do it better, and we need its support going forth.