New AI Model SHOCKED Me
Why It Matters
The model’s real‑time multimodal interaction and translation capabilities could give enterprises a powerful new tool for seamless communication, intensifying competition among AI startups and established players.
Key Takeaways
- •Thinking Machines Labs launches preview AI model with multimodal interaction.
- •Model can see, hear, and interrupt users in real time.
- •Integrated web search and artifact generation enhance conversational depth.
- •Real‑time language translation demonstrated via live friend‑triggered prompts.
- •Founder Mera Marotti, former OpenAI CTO, leads innovative AI venture.
Summary
Thinking Machines Labs unveiled a preview of its next‑generation AI model, marking the first public demo from the startup founded by former OpenAI CTO Mera Marotti. The announcement follows Marotti’s departure from OpenAI and positions the new firm as a direct competitor in the generative‑AI space.
The model distinguishes itself with multimodal capabilities: it can see video, hear audio, and respond with context‑aware interruptions. Integrated web‑search and artifact generation allow it to pull up up‑to‑date information and produce structured outputs, while a built‑in real‑time translator switches languages on the fly.
During the demo, the presenter instructed the system to say “friend” whenever a guest entered the frame, prompting the AI to announce each arrival and instantly translate the speaker’s words. The host described the experience as “something novel in the AI world” that he hadn’t seen in a while.
If the preview lives up to its promise, the technology could reshape customer‑service bots, virtual assistants, and cross‑border collaboration tools, accelerating the race for truly conversational, multimodal AI platforms.
Comments
Want to join the conversation?
Loading comments...