Day 6 - Workgroup Updates, Industrial Experience Panel (Andre van Schaik, Tobi Delbruck, Eric Ryu, Christoph Posch, Christian Mayr), Ethics (Sara Solla) and Social Dinner

The evening before, long after the last sessions had ended, the workshop “disco” was still full of activity. Some groups were debugging hardware, others were still coding, discussing, or trying to get some more results working before the morning updates. Laptops stayed open for at least a few groups until around 2 AM.

A few hours later, somehow, everyone made it back for breakfast and the morning session under another day of clear skies and bright sun.

The morning began with the workgroup updates.

After several days of discussions, experiments, and some late evening/night-work sessions, the different project groups now gave a short overview of where they currently stood. What had worked so far, what had not, what had changed since the original project pitches, and what they planned to focus on during the second week.

Some groups presented first prototypes or simulation results, software pipelines, experimental setups, or revised research directions that had emerged during the week.

The update session gave a useful snapshot of what it looks like when people with different backgrounds come together.

By late morning, the focus shifted from research projects to a different question:

What happens when an idea leaves the lab?

Between Startups, Science, and Industry

The Industrial Experience Panel brought together researchers who had taken very different paths after academia. Among them were Andre van Schaik, Tobi Delbruck, Christoph Posch, and Eric Ryu, who had experience both in Academia but also in building companies, commercializing technologies, or working closely with industry.

The session was intentionally interactive and started with a simple show of hands: who in the audience could imagine working in industry, joining a startup, or even starting their own company one day?

Following that, the ideas about what a successful transition into industry is, and what the goal of entering the market with your product should be, led into a much broader discussion.

Should you build a company, license your IP, join an existing company, or simply continue doing research? And when is a technology mature enough to leave academia?

And just as importantly, why would you choose one path over another?

The panelists shared very different stories. Some had gone through the startup route, others through licensing or industrial collaborations. There were success stories, but also examples of prototypes that never found a market, partnerships that did not work out, and technologies that took much longer than expected to become useful outside the lab.

A recurring theme was that building good technology is only one part of the process. Finding actual users, or customers, turned out to be equally important. At one point, one of the panelists summarized the stereotypical entrepreneur's priority as:

“Customers, customers, customers.”

The discussion also touched on patents, technology transfer offices, early funding, and what several speakers described as the “valley of death”, the difficult phase between a working research prototype and something reliable enough for real-world deployment.

One aspect of the session was that nobody tried to present entrepreneurship as a straightforward path. Several speakers openly discussed mistakes, unexpected pivots, and things they would likely do differently if starting again.

While much of today’s AI discussion is dominated by large-scale models and increasingly ambitious claims, the examples discussed during the panel were much more grounded in the neuromorphic community: event-based sensors, embedded systems, low-power hardware, specialized chips, and applications where efficiency matters at least as much as raw scale.

This eventually led to a broader discussion about scaling itself.

Scaling… and Why

What started as a practical conversation about how to grow a technology gradually turned into a more fundamental discussion.

If you want a technology to have impact, does it need to scale? Does commercialization automatically imply growth? And who decides what kind of growth is actually desirable?

At this point, Sara Solla pushed the discussion in a different direction by questioning some of the assumptions around scaling and technological progress, facilitating a discussion about responsibility. If we as researchers develop technologies that may eventually be used in surveillance, defense, or large-scale infrastructure, when should those questions be discussed? Only once a product exists, or already during the research itself?

The discussion lead into agreement that these questions deserved their own discussion later in the day.

Ethics After Lunch

In the afternoon, an informal ethics discussion group formed.

The conversation moved through a range of topics: regulation, the notion of local societal rules instead of just bottom-up regulation, open-ended curiosity-driven research, and the responsibility of principal investigators to actively discuss ethical implications with their groups, not only once technologies are mature, but already during early research.

Dinner, Awards, and Alghero

In the evening, the workshop moved from the lecture hall to a local Agriturismo in the countryside near CapoCaccia.

Before dinner, the group gathered for the presentation of this year’s Mahowald-Mead Prize, honoring contributions to neuromorphic engineering and the legacy of pioneers such as Misha Mahowald and Carver Mead. This year’s awards went to Mohamadreza Zolfagharinejad from the University of Twente and to the SpiNNaker Project creators.

 

After that, the evening shifted toward good food and conversations, and later, for the people who hopped on the Part-Bus, a trip into Alghero City. They returned around 330am and recovered Sunday morning as best and quickly as they could.