The Wizard of Open Weights: Chris Alexiuk Makes AI Easy With Open Models

Chris Alexiuk, a Toronto-based senior product research engineer in deep learning at NVIDIA, specializes in making open large language models practical for working developers. His rule is simple: if you believe in open source AI, be loud about it. "We should be shouting about this technology from the rooftops," he said in NVIDIA Blog's report. That rooftop is YouTube, where he has built a following teaching curious thousands how to build, fine-tune, evaluate and ship production LLM applications and agents.

This piece is for those working developers and for companies choosing a model for a job. It covers his conclusion that fully open source AI requires the weights, the data that went into a model, and the training recipes. It then follows how he sorts the NVIDIA Nemotron family, from a model like Nemotron 3.5 Lightning to Nemotron Ultra.

Wizard of: what actually changed

After years of teaching developers to build with open models, Alexiuk has concluded that transparency is the precondition for good AI. "Fully open source AI is necessary for the democratization of intelligence," he said. "You should have access to the weights, the data that went into it, the training recipes, everything." On his LinkedIn profile he styles himself "No. 1 Nemotron Fan." That name refers to the NVIDIA Nemotron family of open source, multimodal models built for long-running, self-evolving and specialized AI agents.

Across social media he also uses the pseudonym "Joe Nemotron." "I think Nemotron is the most important project that exists," Alexiuk said. "I'm a firm believer that openness is a very important thing to undertake, so I'm going to be the No. 1 fan of whoever is doing that the best. And that is Nemotron." The LinkedIn title and the pseudonym are how he says that choice out loud: he will be the loudest fan of the open model family he thinks is doing openness best.

Wizard of: how it works

On X he goes by the handle llm_wizard. His avatar is a laughing cartoon caricature sporting a slouchy sorcerer's hat and brandishing a magic wand. He likens Nemotron to a magical library, and its users to fellow wizards in search of knowledge that can move mountains. "You're a wizard in search of prepared spells, and Nemotron has a spell for everything," he said. "Just choose the one you need." Matter-of-fact directness and affable approachability are what make that teaching style appealing.

In non-wizard terms, selection is a category match. "There's a category of model that's going to be close to the kind of work you want to accomplish," Alexiuk said. Nemotron, he explains, supports that whole ecosystem of AI users. An individual developer can use a model like Nemotron 3.5 Lightning for repetitive, high-frequency execution work. A company can deploy Nemotron Ultra for long-running autonomous agents, deep research and enterprise tasks. The prepared spell is the model whose category sits closest to the work.

Wizard of: why it matters

"I think the people who will most help us understand how to build and apply AI likely haven't even heard of it yet," he said. "Trying to get knowledge and information out to as many people as humanly possible is what continues to drive me." YouTube is where that drive shows up, in front of curious thousands learning to build, fine-tune, evaluate and ship production LLM applications and agents. The bundle he wants them to have is the weights, the data that went into the model, and the training recipes.

"I think that openness helps more people understand AI and hopefully leads to more people doing really ingenious things with these tools at the limit," Alexiuk said. "If a tool is amplifying skill, it's amplifying human expression." NVIDIA's profile shows NVIDIAN Chris Alexiuk at left, engaging with developers during the Build-a-Claw event at GTC Berlin. Those tools, for him, are the open NVIDIA Nemotron models he calls the most important project that exists. Curious thousands on YouTube are the people he wants using them.

Wizard of: who is affected

The ecosystem he describes has two working audiences. Individual developers doing repetitive, high-frequency execution work are matched to a model like Nemotron 3.5 Lightning. A company deploying Nemotron Ultra for long-running autonomous agents, deep research and enterprise tasks is the other. Both use the NVIDIA Nemotron family of open source, multimodal models built for long-running, self-evolving and specialized AI agents. Alexiuk's job is the practical layer: he is NVIDIA's Toronto-based senior product research engineer in deep learning for open LLMs.

His path started in a computational physics class at Trent University in Ontario, Canada, where machine learning caught him. "After that, I learned about TensorFlow and that you could build image classifiers that ran on your GPU," he recalled. The initial one-off project alerted him whenever his friend's distinctive pet bird, a Yellow-Faced Gray Budgerigar, appeared in frame on his webcam during shared video game sessions. "Because we were playing video games, I wasn't focused on my friend's webcam, but I didn't want to miss seeing the bird when it appeared," Alexiuk said.

Wizard of: what to watch

"And then my whole life changed, because in that instance I realized it's crazy what you could do with technology," Alexiuk said. Since then, he has been embedding himself in the open source AI community. "To imagine within the span of a few years going from freaking out because I could know my friend's bird was in the camera frame to being able to make any website in a matter of minutes with a model that's running on my laptop — I cannot wait to see what's next," he said.

"It's never been more accessible to get into and learn about AI," Alexiuk said. If another LLM wizard arises to challenge him for the title of No. 1 Nemotron Fan, he said, "It would be the greatest day of my life to be surpassed by an even bigger fan of Nemotron, who had heard about Nemotron through the work I was doing. That would be a capstone to my career that I can only dream of." The NVIDIA Nemotron Labs Livestreams series on YouTube is where he continues that teaching.