Technica Radar 2024

More than AI: These are the key IT trends for 2024

Security, data strategies and artificial intelligence: These three topics rank highly on bbv’s Technica Radar for 2024. In this interview, Marco Ravicini and Patrick Labud from the CTO Board discuss “AI-as-a-colleague”, theft of customer data and the impact of the EU’s Cyber Resilience Act on Swiss SMEs.

06.06.2024Text: tnt-graphics0 Comments
bbv Technica Radar

bbv’s Technica Radar has identified at least ten IT and software trends each year to date. This year, the tech landscape is dominated by artificial intelligence. Should companies focus primarily on AI?
Patrick Labud: There is such hype around AI that many other topics and trends fade into the background as a result. In our Technica Radar, we show that 2024 will be about more than just artificial intelligence. All topics that we have highlighted on our Software Development Quality Map and in the Technica Radar over the years remain valid and important. But: Whether hype or not, most companies will have to contend with AI. Similarly with the topics of data strategies and management and security, the two other major topics on this year’s Technica Radar.

What is the impact of AI on software development?
Marco Ravicini: The hope is that AI will make software development more efficient. It seems at present that AI is good at generating small blocks of code with sufficient information. However, AI is not as successful when it comes to entire classes or complicated functions.
Patrick Labud: The errors that AI generates are not obvious and troubleshooting is often difficult. The key therefore is to ensure that AI-generated code is checked by a software engineer.

Man as the last resort?
Marco Ravicini: Precisely: AI generates very good decision templates based on a set of data. But allowing AI to make decisions on its own at this point in time is not something I would do.

 

bbv Technica Radar
bbv’s CTO board publishes the Technica Radar annually: Marco Ravicini (on left), Patrick Labud and Axel Hohnberg (not in the picture).

How exactly does AI simplify software development?
Marco Ravicini: It helps me, for example, if I am programming in a language I am not very familiar with or in a new ecosystem. This means I am up and running faster in the new environment. But there’s no avoiding editing and checking the code.
Patrick Labud: Thanks to AI, developers don’t need to read as many instructions. There’s no longer any need for me to learn commands in individual cases. Instead, they can be requested and generated verbally or in writing using generative AI. We’re talking here primarily about the level of individual code sections, not about the complex development itself. Despite AI, this still remains the preserve of humans for the time being.

 

“AI is like any technological innovation: It can, but does not necessarily, make you more efficient.”

Marco Ravicini

 

So AI tools are valuable, but cannot provide fully-fledged support?
Patrick Labud: That’s right, because AI is based on statistical relationships and not on logic. If you give it a task, it’s like instructing junior engineers: They can take on tasks themselves within a certain framework, but you still have to keep an eye on them. Artificial intelligence is like a colleague who can help out but doesn’t have decision-making authority or a say – “AI-as-a-colleague”, as it were.

That sounds good. But does AI actually make employees more efficient?
Marco Ravicini: There’s a study that looked at the productivity of customer service employees. It showed that particularly productive employees are slower and the quality of their work diminishes when they work with AI. Innovation and motivation therefore suffer as a result. AI is like any technological innovation: It can, but does not necessarily, make you more efficient. Every engineer must and should decide for themselves which AI tools they want to use for development.

Let’s move on to the topic of security, the second major topic on the Technica Radar 2024. What’s new?
Marco Ravicini: New laws and regulations in the area of IT security are coming into force this year. The EU is set to introduce a reporting requirement soon for hacker attacks and IT vulnerabilities with its Cyber Resilience Act (CRA). Vulnerability management is now absolutely necessary. As a result of the CRA, “Security by Design” must be adhered to throughout the entire software development process. This also affects Swiss companies that have customers or partners in the EU.

What other current developments are there in the area of​security?
Patrick Labud: The Software Bill of Materials (SBOM) has become more relevant as part of “Security by Design”. Industries like medical technology have long since been familiar with SBOM, but now other sectors are having to deal with it, too. The SBOM helps to check software and software components for known vulnerabilities, problems or critical owners. We can see that companies still have a lot of work to do here. Security is still viewed as a cost factor and not as an asset that inspires trust.

Can AI help to ward off cyberattacks?
Marco Ravicini: Yes, it’s conceivable that AI can offer valuable support in finding patterns and anomalies in a system, for example in runtime behaviour.

 

“Customer data that is not collected also cannot be stolen.”

Marco Ravicini

 

The third major topic for the Technica Radar is data management and data strategy. A perennial topic?
Patrick Labud: The topic of data strategy is very topical, especially in relation to AI. Companies today generate enormous amounts of data, but only utilise a fraction of its potential. A data strategy helps to collect high-quality data, combine it intelligently – and ultimately monetise it.

What should companies look out for?
Patrick Labud: A company must basically decide what its goal is with the data. Deciding what to do with the data forms the basis for the strategy and methods to be used to collect and manage data and make it available productively.  Many SMEs find themselves drowning and losing themselves in their data because they maintain it poorly and do not practise adequate data management. AI and Large Language Models (LLM) can help here, for example, to systematically enter data and information in a company database.

So collecting data in advance has therefore had its day?
Marco Ravicini: Data involves time and money. With sensor-based IoT systems, it’s quite possible that preventively collected data could be used subsequently for new use cases. However, this approach is expensive and risky if customer and usage data is aggregated. A lot of resources are needed because such aggregated data causes increased responsibility:  It has to be maintained, protected and kept up to date at all times. Customer data that is not collected also cannot be stolen – and does not involve any time or money.

Can we expect any other major topics for 2024?
Patrick Labud: Certainly the no-code and low-code movement, which not only affects IT employees, but also many other teams. There is also a focus on resilient systems and platform engineering.

The expert

Patrick Labud

Patrick Labud is a consultant and member of bbv’s CTO board. The qualified computer scientist has many years of expertise in usability, user and customer experience (UX/CX) and design thinking.  Today he works primarily as a consultant and speaker for human-centred digital product development.

The expert

Marco Ravicini

Marco Ravicini is a software architect and member of bbv’s CTO board. The qualified automation engineer studied computer science and has an MAS in Human Computer Interaction Design. He is a passionate advocate of the software crafting movement.

People at bbv

“Good ideas are our raw material”

Agile Software Development
Mehr Effizienz und Automatisierung dank Generative AI

How AI agents are revolutionising our working world

AI
People at bbv

«Family and technology are my life force»

Innovation

Comment on article

Attention!

Sorry, so far we got only content in German for this section.

Attention!

Sorry, so far we got only content in German for this section.