Skip to content
UniqueCapital

UNIQUE RESEARCH / ENGLISH ARTICLE

Meta CEO Mark Zuckerberg & NVIDIA CEO Jensen Huang: Everyone Will Have an AI Assistant

Unique Research · Insight · August 2, 2024 · SIGGRAPH 2024 fireside

At SIGGRAPH 2024, Nvidia CEO Jensen Huang and Meta CEO Mark Zuckerberg talked open-source AI and the path to ubiquitous assistants.

Zuckerberg opened with AI Studio — create, share, and discover AI characters, integrated into WhatsApp and Instagram so creators and SMBs can ship bots on Llama 3.1 that auto-answer queries or extend a creator’s reach.

“Every restaurant, every website, may have these AI tools…” Huang said. Zuckerberg: “…just as every business has an email, a website, and a social account, I think in the future every business will have an AI assistant.”

Huang praised Meta’s AI stack — computer vision, LMs, realtime translation — and noted “we all use PyTorch, and it comes from Meta,” calling Meta a leader others only recently noticed.

Why open source

Both stressed open platforms. Llama 3.1’s largest variant: 405B parameters, trained on 16,000+ Nvidia H100s. Zuckerberg: unify modalities under one model as scale rises — someday Facebook/Instagram services may run on a single model. Llama 3.1, he wrote, may mark an open-source tipping point for richer tools, digital avatars, and virtual worlds — and productivity if everyone has one or more digital assistants.

Huang’s vignette: typing in WhatsApp while images generate live and update as wording changes.

Looking ahead

Zuckerberg: AI + real world via glasses (Luxottica partnership) for education, entertainment, work. Huang: interaction moves past turn-based text — models that consider option trees and simulate outcomes. The chat closed with leather-jacket gifts and jokes (Zuck’s hooded fur jacket; Huang’s two-hour-old SIGGRAPH jacket from wife Lori).

Editorial English edition by Unique Research, published on UniqueCapital on August 2, 2024. This translation is based on an earlier Chinese article; the Chinese source is credited separately and is not treated as a line-by-line mirror.

View the Chinese source article ↗
← Back to English research