Why Multimodal Input Matters for AGI: DeepSeek & Anthropic's Approach
dotey · x · 2026-08-02
Countering the claim that multimodality is irrelevant to AGI, the author argues that since human-built physical and digital worlds rely heavily on specific modalities like vision, AI must process these inputs to accurately represent the world and execute long-horizon tasks.
Regarding multimodal output, the author notes that both DeepSeek and Anthropic treat it as a vertical engineering problem (e.g., diffusion models) that can eventually be externalized and called upon by general intelligence. Furthermore, DeepSeek hasn't abandoned vision; its non-open-weight versions already process visual signals, and enterprises are actively integrating ViT modules with it.
More from AGI Musings
- DHH: Starship, Starlink and Grok should have been Europe's Sputnik moments — georgemillo · 2026-10-16
- Hinton hits back at Jensen Huang: 'He has good reason to say AI isn't dangerous — he sells Nvidia chips' — BlackHC · 2026-10-16
- Yudkowsky blasts new interpretability finding as primitive; critic pushes back on doomer logic — AlexTensor · 2026-10-16
- Beff Jezos declares 'civilizational inflection point' in the acceleration of technological progress — beffjezos · 2026-10-16
- Data scientist: pivoting to AI engineering is the safest career bet for the next 5 years — mdancho84 · 2026-10-16
- AI will be superhuman at science regardless, and Europe can't opt out — teortaxesTex · 2026-10-16