Is IT a source of resilience or fragility?
In the face of uncertainty, businesses have made resilience a strategic priority. Although algorithmic predictions are of little help, IT is doubly concerned. Because the resilience of the organization largely depends on its resilience, and because IT must ensure that digitalization strengthens rather than harms the improvisation capabilities of the company’s employees during the next emergency situation.
Consultant at Eraenos, Pierre-Yves Niederhauser led a workshop on antifragility for the French-speaking CIOs of the Digital Circle at the end of 2023. In an interview, he returns to this notion and what it can bring to the way we think about IT management.
A Dutch researcher specializing in cybersecurity and enterprise architecture, Edzo Botjes is passionate about the issue of anti-fragility. While in Switzerland to speak at the 10th Swiss Cyber Storm conference, he answered questions from ICTjournal.
What are the uses and challenges of AI in health?
Breast cancer screening is a promising area for AI where this technology is beginning to establish itself. ICTjournal met three radiologists working in French-speaking institutions, who use software based on deep learning to identify malignant lesions in mammograms. Their testimonies reveal what happens when a specialist works on a daily basis with an AI assistant, between complementarity, complicity and competition.
Solutions based on generative AI are arriving among doctors. Swiss technologies excel in this area and integrate transcription and data extraction tools, in particular to automate the drafting of consultation summary reports and the entry of information into patient files.
Professor at the University of Geneva, Olivier Michielin explained, during a recent inaugural lesson, how AI is transforming oncology, promising tailor-made treatments based on a wealth of biomedical data, including from high resolution imaging.
What tasks for SOCs? What tools? What measures?
A recent report from the SANS Institute shows that alert management dominates activities supported by security operations centers (SOCs), while functions like red-teaming and purple-teaming are less common and typically outsourced . On the tool side, EDR/XDR systems are the most popular technologies. Threat hunting and SIEM are also popular, while AI and machine learning are less popular. Finally, to evaluate their effectiveness, SOCs rely on performance indicators such as the rigor of threat eradication and the number of incidents handled.
What if robots learned thanks to AI?
The robots-LLM association is booming. A recent study highlights the crucial role of this technology in the progress of robotics, particularly with a view to improving human-robot interaction and broadening the fields of applications. Like the Zurich start-up Mimic, which develops advanced technologies for robotic arms and hands integrating generative AI. Despite advances in AI, however, robots still struggle to take on domestic tasks.
Do AI data centers consume too much?
To train, operate and deliver GenAI models, cloud giants are investing heavily in their computing capacities and the construction of new data centers. Built for AI, these data centers will need a lot of energy. Electricity needs and bills will rise and carbon neutrality objectives will move away. Tensions will also increase with other uses of energy and water. These challenges stimulate the development of technologies that somewhat mitigate impacts, such as nuclear power, liquid cooling or even green concrete. Some even imagine sending data centers into orbit.
What if AI predicted… the weather?
Weather forecasting has entered a new era, that of data, machine learning and artificial intelligence.
A doctor in physics, Lionel Moret heads a team using machine learning at MétéoSuisse. In an interview with ICTjournal, he explains the potential and challenges of this technology for weather forecasting, in light of current projects and developments.
Alexander Stauch, Head of Marketing at Meteomatics, discusses the challenges businesses face when it comes to weather forecasting. It highlights the model created by this Swiss firm which provides ultra-precise and frequently updated data.
