Tech-reality check in 2025: What is the hype?

A techno-progressive critique

We have to start with AI. Yes, we have made remarkable progress - just before Christmas in 2024 OpenAI released a series of features, some remarkable, others minor. DeepSeek V3 was launched, outperforming the top models at a fraction of the cost. Google launched Gemini 2.0 to catch up with the top dogs.

However, most implementations still struggle with basic usability issues. I have seen countless companies invest in AI solutions without clear use cases, chasing technology for technology's sake. This is not progress - it is expensive experimentation. Expensive is the keyword here, while AI is easily accessible and used by kids as well as grandparents, the underlying costs that

AR and VR present similar challenges. While Apple's Vision Pro demonstrates impressive technical capabilities, its $3,499 price tag and limited practical applications highlight a crucial gap between technological capability and meaningful utility. The metaverse, or spatial computing as Apple likes to call it, is still searching for its killer app. Research consistently shows that successful technology adoption depends more on solving real problems than on technical sophistication.

Techno-progressive beliefs

The reality is that most organizations still struggle with basic digital transformation. Adding AI, AR, or VR without addressing fundamental operational issues is like putting a Ferrari engine in a bicycle - impressive but impractical. This goes against the basic principle of techno-progressivism: Technological advancement is crucial for progress.

This relates to another fascinating aspect of techno-progressivism: the idea that technological progress should serve as an enhancement for humanity - that is humans can and should be enhanced through technology. Yet, we see emerging technologies remaining exclusive to wealthy early adopters or large corporations. The democratization of technology stops at basic consumer applications while transformative technologies remain inaccessible to most.

Societal & ethical dimensions

Next, perhaps most concerning is how far we are from democratic access to transformative technologies. High-end AR and VR remain exclusive to wealthy early adopters or large corporations. While basic AI tools are accessible through smartphones, professional applications often stay locked behind expensive enterprise solutions. The democratization of technology stops at basic consumer applications while truly transformative technologies - those that could enhance human capabilities - remain inaccessible to most.

The fact that technological progress should serve social justice seems forgotten in 2025. Instead of reducing inequalities, current implementation patterns often amplify them. Non-users of technology will be left behind. Acess remains to few. Take AI in healthcare - while it improves diagnosis accuracy in wealthy urban centers, rural and low-income communities fall further behind. The promise of technology as an equalizer remains unfulfilled, and we see tech-focus companies growing faster than others (see “The Magnificient 7”), widening the gap.

Implementation gap

The reality of tech progress in 2025 is fascinating, but not in a good way: We have created incredibly powerful technologies that we cannot properly implement or validate. It is like building a rocket ship without thinking about where we want to fly - impressive, but ultimately not that useful (maybe Elon Musk is already working on this).

Public science funding cuts have left a vacuum that companies are rushing to fill - primarily because there is a pot of gold involved and nobody that stops them. But here is the problem: While companies push out new AI models every few weeks, scientific validation lags months or years behind - and the gap widens as your read this. And the environmental costs? Training a single large language model uses more energy than hundreds of households annually. We are essentially running a massive experiment without proper controls or consideration of the side effects.

The integration challenge is even more intriguing. We have AI models that can write poetry and solve complex math problems, but struggle to integrate them into basic workflows. In Europe, regulatory uncertainty adds another layer of complexity. We end up with powerful tools that sit isolated from the systems that create value.

This capability-over-utility focus needs to shift. We need fewer impressive demonstrations and more practical solutions that fit into existing workflows and systems. Otherwise, we risk building ever more powerful tools that create more problems than they solve.

The way forward? Focus on integration and validation first, capabilities second. But that would require a fundamental shift in how we approach technological development. And I am not sure we are ready for that conversation yet.