MacroVoices · Thursday, September 10, 2026
Matt Barrie discussed the adoption of NVIDIA DGX Sparks for on-premise AI computing, noting their ease of setup and ability to run various AI models, including open-source options like GLM 5.3. He mentioned that a previous free offering, Ox Alpha, saw dramatic improvements due to extensive training data, raising questions about user data usage.
“So I caught up my local computer shop and I said, um, can I get, uh, a pair of DGX Sparks? And this is the NVIDIA little, um, AI supercomputer.”
“And what's amazing is you can actually load up something like the Pyg coding harness, which is PI. And you can just tell it, please set up the box and, you know, download DeepC or download GLM 5.3 and configure yourself and you sit back and you watch it for 15, 20 minutes.”
“Now, the problem with all of this is, they're all training on it. So if you're kind of wondering why, you know, why it's free, it's because you're probably the product.”