"True game changers fulfill three criteria simultaneously"

Written by Alexander Lorber, on Springer Professional

Read the full article on Springer Professional: https://www.springerprofessional.de/transformation/-echte-gamechanger-erfuellen-drei-kriterien-gleichzeitig-/52194332

‍

In the interview, Stephan Wirries, General Partner at Ventech, explains what makes innovative technologies game changers, how to recognize short-lived hypes, and why artificial intelligence has become the engine of digitalization.

IT Director: Mr. Wirries, digitalization plays a central role in the future viability of Germany as a business location. From your perspective, which technological trends are key in this context?

Stephan Wirries: Three developments will fundamentally change Germany's economic landscape in the coming years. Firstly, there's Agentic Artificial Intelligence (Agentic AI), the transition from chatbots to autonomous digital employees. In January 2026, Anthropic launched Claude Cowork, an AI-powered agent that independently reads and processes files and handles multi-stage workflows – complemented by eleven specialized open-source plugins for sales, legal, finance, and other business functions. SAP followed suit with Joule agents for finance, human resources (HR), purchasing, and supply chain. Gartner predicts that by the end of 2026, around 40 percent of all enterprise applications will have integrated AI agents – compared to less than 5 percent in 2025. Secondly, a structural shift is underway in Software-as-a-Service (SaaS). The per-seat pricing model is ending the trend of paying for services rendered. If one AI agent takes over the work of three human employees, three licenses are no longer needed. This harms net revenue retention, a key metric that shows how much money software companies earn from existing customers. AI programs operate differently: you pay for each individual request and response. Therefore, new AI software companies are changing their pricing – instead of a fixed monthly fee, customers only pay for what they actually use. According to IDC, by 2028, around 70 percent of providers will be billing this way, based on consumption or results. This means that companies will have to completely reallocate their software spending. Last but not least, digital sovereignty is becoming more critical than ever. With AI agents that directly access local file systems and company data, it's no longer just about storage, but about the algorithmic analysis of sensitive business data, including by US systems. At the same time, new EU regulations are coming into effect: NIS 2 makes CEOs personally responsible for IT security, and from August 2026, the AI ​​Act will require companies to disclose where and how they use AI.

Why do some tech topics become true game-changers, while others are just a short-lived hype?

True game-changers fulfill three criteria simultaneously: The infrastructure is mature, the economic benefits are measurable, and the technology can be used without specialized knowledge. If one of these is missing, it remains just hype – as with blockchain, which fascinated technologically but solved few mass problems better than existing systems. Generative AI fulfills all three criteria. The coding tool Cursor achieved one billion dollars in annual revenue in record time. The proof lies not in visions, but in concrete value creation: A subscription for 20 dollars a month replaces parts of software that cost 200 dollars per user. But the real structural change goes deeper. We are witnessing an interface war: AI agents like Claude Cowork are becoming the universal user interface. Anyone who manages their customer relationship management (CRM) via an AI agent never logs into the actual application again. The SaaS application becomes a headless backend – an expensive database without its own interface. This shifts the competitive landscape from the system of record to the system of action. Whoever controls the action layer – the agent – ​​determines the market. My rule of thumb: If a technology needs the phrase "This will change everything," it's hype. If it generates the phrase "This will save me three hours a week," it's a game-changer.

Why is artificial intelligence (AI) now the be-all and end-all, but was doomed to failure in 2002?

The second AI winter ended in 2002. Expert systems were rule-based, rigid, and fragile. They lacked everything: too little data, too little computing power, and too weak algorithms. The breakthrough only came a few years ago. AI was like a race car without a road, without fuel, and without a driver. What has changed since then can be described as a three-pronged convergence. First, data: The internet, social media, and the Internet of Things (IoT) have generated petabytes of data. In 2002, there was neither Wikipedia in its current form nor the amount of text that modern language models need for training. Second, computing power: GPUs enabled parallel processing on a scale that was unimaginable in 2002. Third, algorithms: The Transformer (2017, Google, "Attention Is All You Need") was the key breakthrough—it solved the sequential bottleneck of earlier architectures and enabled parallel processing with true contextual understanding. Crucially, these three factors don't converge linearly, but exponentially. Costs are falling rapidly – ​​OpenAI reduced the price of one model by 80 percent within two months. At the same time, quality is increasing and application is becoming radically simpler. Claude Cowork is a no-code agent that even departments without an IT background can use. In 2002, every single prerequisite was lacking. By 2026, data, computing power, and algorithms will simultaneously be at a level that enables not only research but also industrial application in everyday business. This convergence occurs every 20 to 30 years – most recently with the internet, and now with artificial intelligence.

Read the full article on Springer Professional: https://www.springerprofessional.de/transformation/-echte-gamechanger-erfuellen-drei-kriterien-gleichzeitig-/52194332