Skip to content

New ways of thinking computer science learned from Heidelberg event

  • by

Abhishek Kumar encourages scholars to use the opportunity to apply for a grant to the Heidelberg Laureate Forum.

– I was surprised to find myself to be the only participant from Finland. It’s a great opportunity for young researchers. As a young researcher, one can apply to this event twice at different career stages, for example once as a PhD student and once as a postdoc, says Kumar.

Kumar has worked on security of complex distributed systems and machine learning for constrained computing devices. He is currently a 6G Flagship Researcher at the University of Oulu but splits his time also as a visiting scholar at the University of Helsinki where he works in Sasu Tarkoma´s NODES group. He was previously a doctoral researcher at the Systems and Media Lab, led by Professor Pan Hui.

The experience of the Heidelberg Laureate Forum was eye-opening for him.

– You get to meet great people, and witness first-hand the pinnacle of knowledge. For instance, I got to talk to many Alan Turing award winners, Jack Dongarra, the pioneer of high-performance computing and Supercomputing, Adi Shamir, the creator of the encryption algorithm, and Vincent Cerf who is called the father of the internet. He designed the TCP/IP protocols and the internet architecture, Kumar says.

Computer science needs to be thought of in a new way

During the week-long conference, participants of the Laureate Forum shared ideas with some of the “most exceptional mathematicians and computer scientists of their generations”. These included the recipients of some of the field’s most prestigious awards, like the Abel Prize, ACM AM Turing Award, ACM Prize and Fields Medal.

– Leslie Lampert asked everyone at our lunch table, a simple question that was eye-opening. What he asked was simply “What is execution”? How one approaches the answer to that question tells a lot of one’s thinking about the fundamentals of computing, says Kumar.

– Once you’ve recognized this, you can change your state and think around your bias.

Participants and laureates were encouraged for scientific exchange in their interactions. At the same time, 65 new papers were published. Thus, Kumar was in a privileged place to find out about what is on the edge of Computer Science development, what kind of problems are to be solved in the near future, and what could revolutionize the entire field.

– While starting to think a new and creating a new ecosystem, one has to be aware of possible problems in the future. For instance, if privacy and trust had been thought out rigorously back in the day, we wouldn’t have a lot of these issues today. We need to start thinking security from the level of hardware – software integration today. The triangle of trust, accountability and consent should be in the forefront.

Thinking of terms of state has got to change

Kumar says that one of the laureates, Shwetak Patel, Winner of 2018 ACM Prize in Comping, presented his ongoing work towards an alternative of silicon-bases chips, which are the source of much of the embedded carbon. The alternative solution should promote high substantiality and address the issue. The current challenge with this ongoing work is to get high reliability in different conditions, the chips should be able to work out in weather ranging from the arctic to African heat.

– At the same time new supercomputers and ecosystems such as metaverse need more performance. For a high-performance internet would need to rely more on distributed systems of multiple cores and many programs, aka parallelism.

– This would allow many things to come together, both physical and augmented. For new telecommuting styles would need to be created. But the problems are bandwidth and latency. New 3D hologram calls could be needed for certain highly detailed and complex work, such as building architecture of high-rise buildings via remote collaboration.

– In the way of development is getting stuck in old ways of doing. The best lesson I took with me is how I used to think about “state” in computing. We need to free ourselves from these old binaries (or or 1) thinking style. For instance, developing programs for quantum computers could be easier if one forgot everything about traditional bit-based programming beforehand.

Heidelberg event participants are selected by a panel of international reviewers on the basis of their research experience, social engagement skills, and letter of motivation.

Leave a Reply

Your email address will not be published. Required fields are marked *