Every business has different data, workflows and standards. Its AI should reflect that. Watch to see how NVIDIA Nemotron open models help teams build specialized AI they can trust, control and customize, then improve against the outcomes that matter most.

Learn more:

if the data and workflows underneath belong to one provider, customizing the model changes little

The future of enterprise AI isn't one-size-fits-all—it's open and hyper-customized.

makes sense. your data is your moat, your ai should reflect that moat

把ceo裁了换你们的ai吧

@trythreews

Open weights plus your own data is where most enterprises end up. @techimpactTV

What a great guy, indeed knowledge is attractive

Hey @grok what are the comparative stats of non Chinese open models - Nemotron, OSS, Mistral, Gemini…

この内容、日本語で詳しく書きました Wrote a detailed take in Japanese:

Sovereign AI argument in one line. Custom, controllable, on your data.

Lower gpu prices. Msrp should be msrp.

Yall fucking gonna get sued out the ass

Open models letting each business tailor AI to their own standards makes sense.

The real value of an open model at work is not just saving money. It is being able to tune its behaviour around a company’s own rules, data and way of working.

everyone who is pro-AI deserve nothing but burning in hell and their entire families get run over by a truck

IDGAF

NVIDIA is a Taiwan company that is based in America is for security. They need to move their headquarters to Taiwan where they have all their manufacturing done. Pretty scary that America’s allowing a foreign company country to manufacture all their AI.

The point about AI needing to match each business's own data standards is spot on. Customization and control are what make it truly useful.

Specialized AI only works when the underlying behavior stays stable. HELIX provides the stability layer that keeps custom models predictable, controllable, and aligned with the outcomes that matter.

Strong point in the video: enterprise value depends on control over data, behaviour and evaluation, not just benchmark scores. The practical test is improving a model against real outcomes without losing auditability.

This is exactly why NVIDIA keeps pulling ahead. Every company has its own data and workflows and Nemotron finally gives them open models they can actually own, control, and specialize. Jensen Huang doesn’t just talk about the future of AI, he builds the tools that let the entire industry shape it on their own terms. Absolute legend. 🔥

Pedos run this company!: