The full programme is now available and everyone who signed up in the last couple of months will have received an email to register for four parallel sessions. There is a wide range of sessions to choose from covering many different Open Science themes, so make sure to register fast to don’t miss out on your favourites!
Several UM employees are also organising a session. One of these employees is Guangzhi Tang, assistant professor in edge computing at Maastricht University (UM) in the Department of Advanced Computing Sciences (DACS). His research focuses on neuromorphic computing, where he explores ways to make artificial intelligence (AI) more efficient and affordable by drawing inspiration from how the human brain works. Beyond research, he’s devoted to educating the next generation of AI practitioners at DACS. His goal is to ensure they are not only technically proficient but also ethically aware and socially responsible for a sustainable future of AI.
Above: Guangzhi Tang
Improving accessibility
Open Science is a key part of Tang’s work: “I believe that sharing knowledge openly helps create a more collaborative and productive research environment. I have most of my scientific papers openly accessible. Currently, I am building my research on top of openly accessible software and hardware. At the same time, my open-source implementations on GitHub support other researchers in advancing the field. To further improve accessibility, I frequently present at conferences focused on neuromorphic computing and edge AI, where I provide detailed insights into important developments in the field.”
The benefits and risks of combining AI with Open Science
Tang thinks Open Science brings several key benefits to the fields of AI and Neuromorphic Computing: “First, sharing open-source implementations of scientific papers is becoming standard practice in AI and machine learning. This allows researchers to verify results and confidently build upon previous work, speeding up progress in the field. Second, openly available benchmarks and datasets make it easier for researchers to fairly compare their results with others, leading to more reliable findings. Third, access to open hardware lets researchers test their ideas on real neuromorphic systems, ensuring their work is applicable in real-world scenarios. Open Science also enhances education by giving students access to these tools and platforms, making learning more practical and hands-on rather than purely theoretical.”
Tang indicates that AI and Open Science can work together in many positive ways: “For instance, making AI systems more accessible through open platforms can help democratise the technology, which is often dominated by large corporations. Additionally, AI can assist researchers by quickly finding relevant information. However, there are also potential risks. One concern is that AI systems might be biased, meaning they could limit researchers to a narrow view of the field. Another issue is that AI can generate incorrect information, which researchers need to verify before sharing it openly. Finally, AI trained on open data could be misused in unethical ways, so safeguards are needed to prevent harm to society.”
National Open Science Festival
Tang will be hosting a session on Open Neuromorphic Computing at the National Open Science Festival: “Neuromorphic computing, which focuses heavily on hardware, has often been a closed-off field due to intellectual property concerns. While this protects certain innovations, it also slows down progress because many researchers are forced to reinvent the wheel and struggle to access the necessary hardware to test their ideas. As a researcher in this area, I believe it’s important to push for more openness in the field. During this session, I hope we can explore possible solutions and learn from other fields’ experiences with Open Science to drive faster, more collaborative progress.”
Share your story
We are always looking for stories on Open Science that show how you implemented Open Science in your work. These can be both success stories (for example on what you did to make your Science more Open (and FAIR)) or attempts at Open Science that didn’t work out the way you expected (for example because the required infrastructure was lacking) and what you would recommend tackling this problem. Please contact us if we may approach you to share your story with the UM community. Do you have a question in general on Open Science? Don’t hesitate to contact Dennie Hebels, the Open Science officer.
0 Comments