2018 technology trend predictions
There are six technology areas that IT contractors will see come into their own in 2018 but more than that -- they’ll likely influence how you operate as a freelance IT consultant, in both your private and professional life, writes Carl Austin, chief technology officer at IT consultancy BJSS.
Cloud Software Development
In 2018, the trend towards serverless, truly cloud native applications is expected to grow. Inevitably this will have an impact upon the skills demanded by software delivery firms. For example, the demand for sysadmin and cloud infrastructure skills, currently at a high premium, will start to reduce alongside the need to administer infrastructure.
Currently many organisations are often too concerned about cloud provider lock-in that comes with the serverless approach and end up investing in unnecessary complex and expensive solutions. While companies should consider the potential issues of cloud vendor lock-in, they should also look at the significant benefits that it offers.
Security is another area that has been impacted by the increased pace of delivery enabled by the cloud. Security is often seen as slowing down the delivery process, either frustrating teams or worse causing them to forego or work-around it.
Security experts need to adjust to this faster pace by helping all members of a development team embed security practices in their everyday work. In addition, teams should include security tooling in their continuous deployment pipelines. These activities will help embed security into delivery, moving from DevOps to DevSecOps.
My belief is that machine learning and cognitive computing will be the most widely applicable and adopted technological movements we have ever seen. It will drive the 4th industrial revolution and has the potential to be the cause of a social revolution. But when it comes to Artificial Intelligence (AI), businesses still have a long way to go and it doesn’t always start with jumping straight in at the deep end.
Firstly, enterprises need to truly understand the applications machine learning and cognitive computing can offer, so they can identify the areas to apply them to. Those that are already experimenting with AI are progressing into production, but often haven’t thought about the process of industrialisation or how to realise the benefits of implementation.
Companies that are moving from experiment to production need to be more forward-thinking about how they create a 'factory line' -- greater experimentation and a repeatable route to production. Cloud providers are starting to create valuable offerings in this space, including Amazon SageMaker and Azure ML Workbench.
Privacy and GDPR
With the EU’s General Data Protection Regulation set to come into effect in May, it’s no surprise that privacy is on the top of everyone’s agenda. Already companies are looking at how to make existing practices GDPR-compliant, but what about making new things GDPR native? It’s important that GDPR requirements are being built into new software to save time and money.
Companies who wish to benefit from consumers personal data should consider providing benefits in return for ‘opt-in.’ The reality is that most consumers do not value their personal data particularly highly and will often agree to share their data for a small incentive. If enterprises are transparent and up-front, then consumers will be more willing to part with their data.
Once GDPR is implemented, it’ll be interesting to see how far the EU goes in identifying and penalising firms that breach the new regulations. Considering the regulations are still somewhat ambiguous, I wonder how the regulations will stand up to legal challenge.
The surface area for security attacks is growing quickly as intelligent devices, chat interfaces and digitisation continue to advance.
Machine learning and chat interfaces are often black box which means it's hard to tell when they have been compromised and begin providing results that are at the whim of an attacker. This could lead to attacks that go undetected for a long time, so it's vital enterprises learn to secure these systems.
Increasingly DevSecOps is being viewed as the next step in the DevOps journey -- security should be part of the culture of a team, and tooling should support this. It’s therefore important that developers, architects and testers learn to threat model and write security tests, while tools should be used for automated scanning to detect vulnerabilities. With the whole team considering security throughout, the true security ‘experts’ can focus their energy on performing constant penetration testing and simulation.
Growth of Blockchain
When it comes to Blockchain, my fear is that the industry is adding to the hype that is causing over application and technology envy. There is no denying that a small number of enterprise organisations have production Blockchain implementations, but this is not to suggest it is ‘enterprise-ready’.
Currently, enterprise-ready technologies feature an advanced ecosystem of third-party tools and have withstood numerous production implementations -- areas that Blockchain has not yet successfully navigated.
Nevertheless, Blockchain is certainly one to watch out for. Public cloud vendors are taking complexity out of creating and running nodes and networks with PaaS offerings and quick-starts and the tooling ecosystem is being developed.
It is important that enterprises also duly consider less complex technologies for implementation of production solutions. Businesses should be investing in experimentation, up-skilling and proof of concepts to validate the applicability of Blockchain to specific use cases they may have.
Cloud Computing vs. Edge Computing
As technology advances, naturally we will see an increase in intelligent devices running smarter capabilities such as deep learning inference. In many such cases the latency of response is important and thus a round trip to the cloud, often a distance away, is inefficient, driving the implementation of edge computing.
My belief is that supporting edge computing will actually help cloud providers gain greater market share for the things that really are cloud workloads. A good example is Amazon Greengrass which now supports running deep learning inference at the edge, while integrating with its other tools. This means that you are more likely to use Amazon Web Services to do the data processing and training to create the models to then push out to the intelligent devices.
Be in no doubt; these six areas will be majorly evolving, even dominating over the next 12 months -- our IT consultancy will be ready for them. Will yours?