Blog posts on artificial intelligence

Some jobs just aren't a good fit for drones. Tight spaces, ground-level inspections, or environments where flying isn't allowed – that's where a ground-based rover can step in instead. RISE researcher Richa Upadhyay talks about a new rover demonstration that combines autonomous mobility, AI, and edge computing over 5G.

The year is 2036, and electric air travel has become part of an increasingly autonomous ecosystem. It's early morning, and I'm on my way to Helsinki to speak at the Autonomous Systems Forum. The airport is just a few blocks away and feels more like a small train station than the large airports we were used to ten years earlier.

A conversation with Anna Ryberg about AI, competence, and the human role in working life. AI companies are hiring philosophers. Communicators call their AI tools "colleagues." That AI increases productivity is now taken for granted. But what does that actually mean for the human role at work?

A conversation about agentic AI, trust, margins of error, interfaces — and the human ability to navigate complexity

We are moving into a new era of work. AI is becoming an active part of how work gets done. As AI systems become more capable, the skills people need are changing too. The challenge is no longer only to produce the answer, but to define the task, guide the process, assess the quality and take responsibility for the consequences.

The year is 2035. A social worker in a medium-sized Swedish municipality is walking into the last meeting of the day. Her AI agent has already compiled case records, previous decisions and relevant anomalies from three different systems. It has indicated where the information is uncertain and suggested several possible ways forward.

Something is happening with AI—and it’s not what we discussed a year ago. The focus has shifted from what AI can do to what it actually does, where it’s used, and what it demands in practice. It’s a more mature, more important conversation. This year, three shifts define a new chapter for AI.

This Futureframing is written from a forward-looking perspective, set in 2035. I was invited to contribute to this issue of the State of AI newsletter by RISE, sharing my reflections on how AI may reshape cybersecurity over the coming decade — exploring both emerging threats and new forms of resilience.

When AI Is Reshaping Society, Talking About It Is No Longer Enough. A blog post on AI, gender equality, and why an escape room–inspired approach can transform how we learn and engage.

AI development is moving quickly. This is clear through rising investments, more patent applications, and the rapid spread of generative tools in the workplace. At the same time, statistics show that many small and medium-sized enterprises hesitate to take the step. Why are you waiting with AI?

How a needs-based investigation contributed to a lightweight approach to defining digital infrastructure – Lessons learned from an AI-enabled quality inspection use case

In 2035, I sit at home in Göteborg looking for travel ideas with an AI tool built on internationally developed foundation models, delivered by European and Swedish providers. The AI works for me and protects my data. At work, similar systems handle routines and simulations. We work six-hour days, and I volunteer weekly at an elderly care home

The year is 2035. As I sip my morning coffee, a neighborhood robot passes by and pauses in the park to sync its local model with the city’s digital twin: soil moisture, energy flows, upcoming repairs. Life looks normal, yet beneath it runs a quiet network of robots, sensors, and AI agents keeping the city alive.

By 2035, AI will form the backbone of — and be fully integrated across — public digital services and physical infrastructure. Open Source AI will be critical for nation states, industries, and citizens to ensure transparency, control, and sovereignty over models and data inputs, and to avoid ending up under the dominance of any third-party actor.

Multimodal large language models can draft reports, analyze images, and aid decisions, yet they remain surprisingly vulnerable. Tiny, imperceptible adversarial perturbations can trigger drastic errors, yielding nonsensical or harmful outputs and undermining reliable, safe AI deployment.

At a recent meeting of AI ambassadors from across RISE, it became clear that we are in the midst of a system-wide shift. Fifty presentations, fifty voices, one shared insight: AI is becoming a foundational layer in everything we do.

At the Turing–Nordic workshop in Oslo, researchers and policymakers explored how AI can help safeguard the Arctic. The key insight? Real progress doesn’t come from perfect data or tidy solutions, but from embracing the mess—contradictions, failures, and uncertainty that reflect how the real world works.

What does it take to build an AI future that’s fair, robust, and aligned with our values? In this post, Sverker Janson shares insights from a recent futures workshop at RISE, where we explored what it would take for Sweden and Europe to lead the way in shaping a trustworthy AI ecosystem.

Reversed design with AI turned into part experiment part art project in designing a unique stage set for the Computer Science Open House 2025 at RISE. It began as a lighthearted message in a Teams chat and soon evolved into a shared vision.

The AI year 2025 is in full swing. It's clear that the pace isn’t slowing down. Momentum is increasing — new models, concepts, and technologies are emerging almost daily. For those unfamiliar with the field, it can feel overwhelming to keep up, and even those who are well-versed may struggle to maintain an overview of the shifting AI landscape.

In recent years, nature restoration has taken on an increasingly central role in both Swedish and European environmental policy. A clear example is the EU’s new Nature Restoration Regulation, which places greater demands on Member States to restore and monitor natural environments and biodiversity.

OOAKI (one-of-a-kind-living) is a Swedish sofa refurbishment startup that worked with RISE to develop its circular business model. In a conversation with RISE, OOAKI’s Chief Digital Officer, Lennart Svanberg, discusses how this shift was inspired by a closer relationship with customers.

We all depend on digital services to work and be secure, from the smartwatch on our wrist to advanced control systems in industry. Today, cybersecurity is a basic prerequisite for everyday life, as essential as electricity or running water.

Optimal performance of Generative AI necessitates the implementation of liquid cooling! In two preceding blog entries, we delved into the challenges and substantial energy consumption associated with training and utilizing expansive AI models, such as in the case of ChatGPT. In this blog post, our objective is to go into the thermal management.

In a previous blog post, we posed the question, "Does generative AI run on thin air?" and discussed the associated challenges and strategies for mitigating energy use. We even put this question to ChatGPT, and it couldn't provide a clear answer.

In today's digital landscape, technological advancements are reshaping the way we communicate. One of the most groundbreaking developments is deployment of 5G and in the future 6G. These next-generation networks are not just about faster downloads; they have the potential to revolutionize how we experience connectivity in every aspect of our lives.

Have you tried asking ChatGPT how much energy it uses to train itself? It answers “As an AI language model, I don't have direct information about the specific energy use involved in training me."

Let’s take a peek into the future of how much energy data centers in Sweden will use by 2030? The report "Energy Use in Data Centers and Digital System", written by us at RISE data center ICE, was recently released for the Swedish Energy Agency. Here we figure out how much energy that will be used 2025 and 2030!

As a UX Researcher, it is our job to understand the user's needs and expectations for a product or service. This is a critical factor in creating a positive user experience. But understanding the user is not always easy and can require a lot of time and resources.

Can you list some credible arguments why we should not include more people/perspectives when developing new services and solutions? I've really tried to come up with some good arguments but can't.