As I finish the last of the mince pies and try once again to decode the lyrics to Auld Lang Syne, I’m pondering technology and what might be the hot topics in 2023. I’m not thinking about all areas of technology; I’m not going to talk about electric vehicles, renewable energy, gene-editing, humanoid robots or all the other cool stuff, even though that’s all very exciting. I’m just thinking about what’s going to affect our customers and our own ways of working here at Airwalk Reply, in the software, cloud, DevOps, automation and data engineering world. Not everything I mention here is going to become fully formed by the next time we’re singing carols, but we’ll certainly be talking about it and thinking about how we want to use it.
Artificial Intelligence (AI) is going to move from being something ‘out there’ that a bunch of specialists use in niche fields to being sat right here beside us. ‘Generative’ AI, so called because it actually generates new content, can create images, videos, literature, music, lyrics or, most importantly for us, code. AI-powered systems can already correct application code you wrote or just write the code for you in the first place. The number and types of languages it can master are unlimited, if enough training data can be provided. The high-profile example is ChatGPT from OpenAI (who also brought us DALL·E 2, which could create pictures of just about anything that you could describe in a line of text). ChatGPT accepts a question or task in natural language, such as English, and provides answers as text. These answers could be simple statements, essays, poetry, snippets of code or entire programs. You could ask it to write you some code, specifying the programming language and exactly what that code is to do, or you could ask it to debug or optimise your existing code. And ‘code’ isn’t limited to traditional programming languages; it can write infrastructure as code, policy as code, just about anything and to a surprising standard.
Now, I don’t expect companies to start firing software engineers and replacing them with AI just yet, but tools like these are going to become powerful accelerators. Remember that developers have always had tools to assist them – arguably nobody starts with pure hardware and starts typing – but the difference the tools can make when placed in the right hands is going to accelerate exponentially.
Extended Reality, or ‘XR’ is the catchall term for Virtual Reality (VR) and Augmented Reality (AR) as the two are starting to blur. While you may associate these technologies with gaming, virtual tourism or making your TikToks looks cooler, there are plenty of other areas where they’re going to start appearing in our day-to-day lives. The commercial use of XR isn’t just about doctors helping patients thousands of miles away or a customer seeing how nice that new sofa might look in their living room. In an office workplace setting, remote meetings will become less about watching your colleagues sit in their little two-dimensional rectangles neatly arranged on your screen and become more about feeling like you’re sitting alongside them, sharing ideas on the table, not on a nasty virtual whiteboard.
Blockchain, despite some high-profile screwups this year, is not going away. There are still plenty of solid applications for the technology that don’t involve cryptocurrencies (though there are also still plenty of applications that use blockchain but shouldn’t – see the flowchart on page 3 of Do you need a Blockchain? by Wüst and Gervais). The concept of a distributed ledger whose integrity can be verified independently is great for supply chain, healthcare, music rights, and countless other industries. In the public sector, blockchain can be used to verify documents and identities, reduce waste and to ensure money is going to the right places. It can be used to streamline the massively bureaucratic and wasteful process of delivering foreign aid. Within financial services, start-ups and established players alike are still looking at how blockchain could be used to simplify asset tracking, lending, settlements, fraud prevention and payments. Non-fungible tokens, or NFTs, are still finding fans despite a dropoff in popularity in 2022. These types of asset shouldn’t be ignored, particularly in the wealth management sector where they may start to form a small but significant part of their more technophile investors’ assets.
While we’re not all suddenly going to have quantum laptops on our desks in 2023, the pace here is increasing. Huge amounts of money are pouring into research and we’re going to see more and more headlines about breakthroughs in the coming months. Large financial organisations are already running trails allowing them to crunch through massive market predictions or risk models at a pace unheard of with conventional technology. These efforts are often a partnership with one of the major cloud providers, as the likes of Google, Microsoft and Amazon try to secure a competitive edge.
People have been talking about multi-cloud strategies for a few years now. When organisations were just starting out on the road to the cloud and only had a few applications running on AWS, Google or Azure the risk to their business of a failure of their provider was relatively low. A few years have passed and now some organisations have a large percentage of their workloads running in the cloud and they don’t want to leave all their eggs in the one basket, either in terms of provider failure, or in terms of cost at contract renegotiation time. In financial services and public sector, organisations will be forced to think this way by regulation to avoid what is referred to as ‘concentration risk’. Various tools are available to help migrate running applications from one cloud to another and containerisation has certainly made life easier, but the use of these tools is mostly motivated by an operational desire to reduce cost, or to avoid short-term or localised outages at cloud providers.
In 2023 we’re going to see more pressure to plan for an emergency exit from any given provider should the worst happen. And like old-fashioned backups or business continuity procedures, these plans won’t be worth the paper they’re written on unless they’re tested. And it follows that the requirement to test will force applications to be designed to be moved, thereby giving the operational and cost-related multi-cloud benefits I described earlier as a happy side-effect of having your cloud ejector-seat primed and ready.