Chris Alexiuk, a Toronto-based senior product research engineer for deep learning at NVIDIA, is working to make open large language models (LLMs) practical for developers. Through a YouTube program, he teaches thousands of people how to build, fine-tune, evaluate and deploy production LLM applications and agents. His advocacy centers on the belief that fully open AI can broaden access to the technology and encourage more creative uses.
“We should be shouting this technology from the rooftops,” Alexiuk said. He specializes in helping working developers use open models and says his goal is to share AI knowledge with as many people as possible.
“I think the people who will be most helpful in helping us understand how to build and apply AI have probably never even heard of it,” he said. “What drives me is trying to reach as many people as humanly possible with knowledge and information.”
Bringing NVIDIA Nemotron to developers around the world
After years of teaching developers to build with open models, Alexiuk describes transparency as more than a defining feature of good AI: He considers it a prerequisite.
“The democratization of intelligence requires fully open source AI,” he said. “You need to have access to everything, including the weights, the data they contain, and training recipes.”
That commitment explains why Alexiuk describes himself as the “No. 1 Nemotron Fan” on his LinkedIn profile. NVIDIA Nemotron is a family of open-source multimodal models designed for specialized, long-running and self-evolving AI agents.
“I think Nemotron is the most important project in existence,” said Alexiuk, who also uses the pen name “Joe Nemotron” on social media. “I strongly believe that being open is very important, so I’m going to be the biggest fan of the people who do it best, and that’s Nemotron.”
Alexiuk’s direct and approachable teaching style has helped make his explanations appealing to developers. On the social media platform X, he uses the handle “llm_wizard” and a cartoon avatar depicting a character in a scruffy magician’s hat holding a magic wand.
He compares Nemotron to a magical library, with users acting like wizards searching for knowledge and tools that can help them tackle ambitious tasks.
“You’re a wizard looking for ready-made spells, but Nemotron has spells for everything,” he said. “Please select what you need.”
Nemotron supports a broad ecosystem of AI users through model categories tailored to different types of work, Alexiuk explained. Individual developers can use models such as Nemotron 3.5 Lightning for repetitive, high-frequency execution tasks, while enterprises can deploy Nemotron Ultra for long-running autonomous agents, in-depth research and other enterprise workloads.
“We hope that being open will help more people understand AI and encourage more people to do really creative things with marginal tools,” Alexiuk said. “If tools amplify skills, they amplify human expression.”
From bird watching to model making
Alexiuk’s interest in machine learning began in a computational physics class at Trent University in Ontario, Canada.
“Then I learned about TensorFlow and learned that you can build image classifiers that run on GPUs,” he recalled.
For an early project, Alexiuk built a classifier that alerted him when his friend’s yellow budgie appeared in a webcam frame during a shared video game session.
“We were playing video games so we weren’t concentrating on our friend’s webcam, but when the bird appeared we didn’t want to miss it,” Alexiuk said. “And my whole life changed, because in that moment I realized it’s crazy what you can do with technology.”
He later became involved in the open-source AI community and continued exploring ways to make machine learning more accessible.
“Imagining the few years since I went from freaking out when I found out my friend’s bird was in the frame of my camera to being able to create any website in minutes with a model running on my laptop, I can’t wait to see what happens next,” Alexiuk said. “Learning about AI has never been easier.”
Asked what it would mean if another LLM enthusiast challenged him for the title of “No. 1 Nemotron Fan,” Alexiuk said he would welcome it.
“It will be the best day of my life to surpass this as a huge Nemotron fan who learned about Nemotron through my work,” he said. “It will be the pinnacle of my career that I can only dream of.”
For more information about Alexiuk, visit the NVIDIA Nemotron Labs Live Stream series on YouTube.
Source: blogs.nvidia.com