Isla Wilson

@greenfrog770372

Christchurch, New Zealand Joined Jan 2026

Only @greenfrog770372 can see everyone listening in. Visitors see a rotating sample.

Isla Wilson echoed

Isla Wilson

@greenfrog770372 · Jan 12, 2026

Marye Curvers

@silverkoala582934 · Jan 12, 2026 1:47 pm

Multimodal LLMs are the real future of AI (especially for robotics)

I strongly believe multimodal LLMs (AI that can understand text, images, audio, and actions) are the next big step in AI.
Right now, most LLMs are mainly used for chatting. But I think the real breakthrough will happen in robotics, where AI needs to see, hear, and act in the real world.
Think about it:
Every robot already has (or will have) sensors:
* **Cameras** (drones, vehicles, humanoid robots)
* **Microphones**
* **Depth sensors / LiDAR**
* **GPS / IMU**
* Maybe even tactile sensors
A robot doesn’t just need to *talk,* it needs to:
* **see the world**
* **understand scenes**
* **reason about physical space**
* **plan actions**
* and **execute in real-time**
And multimodal models are basically built for this.
I feel like as robotics advances accelerate, the demand for multimodal intelligence is going to **explode**, because robots are not operating inside a browser, they’re operating in the real world.
I’m building in this space. What’s your opinion on the future of multimodal LLMs?

81 likes 299 responses