Technology conferences offer a chance to step away from daily project work and look at the bigger picture. What is changing in software development? Which ideas deserve closer attention? And what can we learn from the people building tools that millions of others rely on?
Zarego’s visit to Nerdearla 2026 brought those questions together. With a program spanning artificial intelligence, open source, infrastructure, security, design, and science, the event provided plenty of reasons to listen, exchange perspectives, and keep learning.
We also interviewed Roger Dingledine, co-founder of the Tor Project. Our conversation explored privacy, decentralization, the community behind Tor, and AI’s emerging contribution to security research.
His perspective connected with a recurring theme: technology depends on the people who build, maintain, explain, and use it.
Five Days of Shared Curiosity
Nerdearla Argentina 2026 took place from September 22 to 26. It began online, continued with workshops at Santander, and moved to Ciudad Cultural Konex for three days of talks and activities, accompanied by streaming.
The agenda covered development, AI, infrastructure, security, UX, product, management, and open source, alongside science and activities for children. That variety gave the event its character. Discussions about agent architecture shared the program with scientific discovery, product decisions, and robotics.
For Zarego, this breadth is valuable. Software projects combine technical constraints, business objectives, and human expectations. Exploring different disciplines helps us ask better questions and approach familiar problems from another angle.
The People Behind Everyday Software
The lineup included figures whose work helped shape modern computing: Rob Pike, co-creator of Go and UTF-8; Monty Widenius, creator of MySQL and MariaDB; D. Richard Hipp, creator of SQLite; Jean-Baptiste Kempf from VideoLAN; and Roger Dingledine from the Tor Project.
Their participation brought a perspective extending beyond the next release or emerging framework. Widely used software must survive changing requirements, growing communities, and years of maintenance.
The panel “The Philosophy and Power of Open Source Leadership” brought together Hipp, Widenius, Kempf, Dingledine, Robin Bender Ginn, and Madelyn Olson.
Its subject highlighted the organizational work behind lasting technical projects. Architecture matters, and so do shared ownership, communication, documentation, and helping others contribute.
AI and the Changing Work of Engineers
AI occupied a prominent place in the agenda, with sessions examining how developers can build, integrate, and evaluate increasingly capable systems.
Nate Barbettini’s “Model Context Protocol in Plain English” addressed MCP, while Carlos Gauto’s workshop on persistent agent memory explored continuity across interactions. These subjects connect with practical challenges surrounding AI applications: accessing tools, retrieving relevant information, and retaining useful context.
Merve Noyan’s “Local AI Ecosystem” explored local inference. Omar Sanseviero examined multimodal agents, generative media, and open models. Together, these topics illustrated the expanding choices available to developers.
Kent C. Dodds’s “The Last Software Engineer” approached the discussion through engineering judgment. His session emphasized accountability, constraints, evaluation, and deciding what should be built as AI takes on more implementation work.
For us, greater capability makes thoughtful decisions more valuable. Teams still need to understand the problem, choose appropriate tools, and determine whether the resulting software serves its purpose.
Our Interview with Roger Dingledine
During the event, Zarego interviewed Roger Dingledine, co-founder and original developer of the Tor Project, which develops free and open source software to protect people from tracking, censorship, and surveillance.
We began by asking about Argentina. Roger explained that this was his second visit, following a previous trip that included technical events and time in Patagonia.
He appreciated gatherings where people care about improving the world, describing events like Nerdearla as opportunities to discuss nonprofits and free software. He also noted the enthusiasm of people eager to learn, while being careful about drawing broad conclusions from his limited time at the event.
When we asked whether he had imagined Tor reaching its current scale, his answer was direct: he had not. It began with a personal goal of being safer when visiting websites. Other people wanted to use it, and its purpose grew.
Roger also explained why diversity among users matters. A privacy network needs people with different backgrounds and reasons for participating. If everyone using it were a computer scientist, that would already reveal something about each user.
Decentralization Needs a Community
Asked which architectural decision had helped Tor endure, Roger first pointed to decentralization. He then offered a broader answer: community building.
People worldwide contribute resources because they want to help. But his explanation extended beyond running relays. Learning how to explain privacy to friends and family is also a contribution.
That observation connected directly with Nerdearla. Technical communities spread knowledge through everyday conversations as well as code. Helping someone understand privacy supports the same mission as developing tools that protect it.
Roger also touched on Tor’s nonprofit structure and the continuing need to raise funds. His comments highlighted a challenge facing public-interest technology: sustaining an organization and its mission requires ongoing effort.
For us, the interview offered a wider view of technical sustainability. A durable system needs architecture, resources, and people who understand why it should continue to exist.
AI’s Contribution to Security Research
We also asked Roger about AI tools. He said he mostly does not use them himself, but identified a recent benefit for Tor: researchers using AI had submitted useful security bug reports.
He welcomed those contributions and described the next step as fixing known issues and seeing what researchers discover afterward.
It was a concrete example of AI contributing to an established engineering process. Its value came from identifying actionable problems that maintainers could investigate and address.
The security program explored related challenges. Germán Ariel Borda’s session on threat modeling for AI applications proposed a practical method using a customer-service chatbot as an example.
Our takeaway is to evaluate AI through specific outcomes and responsibilities. Finding vulnerabilities, reviewing evidence, and implementing fixes remain connected parts of the work.
Bringing the Ideas Back to Zarego
Nerdearla also encouraged curiosity beyond professional software development. Activities included an innovation workshop using LEGO materials, children’s robotics, Daniela Di Bella’s session on brain development, and Lucas Corsino’s exploration of Formula 1 telemetry.
These formats reflected the value of learning through experimentation and unfamiliar perspectives.
For Zarego, visiting Nerdearla was an opportunity to connect those perspectives with our work. The talks raised questions about engineering, AI, and security. Our conversation with Roger added a reminder about the people and shared purpose that sustain technology.
We bring those questions into our team discussions: how to build responsibly, share what we learn, and make software useful for the people who depend on it.
Let’s talk.
