Thinking Machines Lab details interaction models, which can think and respond in real time, letting users and AI interact continuously for better collaboration
Today, we're announcing a research preview of interaction models: models that handle interaction natively rather than through external scaffolding.
Thinking Machines Lab
Context & Ripple Effects
This preview establishes an interaction-focused research track at Thinking Machines Lab: the company is positioning AI around continuous collaboration rather than a sequence of isolated prompts and responses.
Subsequent coverage sharpens that direction. The lab describes its mission as making AI shapeable by people and organizations, and later introduced Inkling as a broad-capability open-weight MoE model—two complementary routes to user control and adaptable general-purpose systems.
First-order effects
Thinking Machines Lab now has a research-preview vehicle for testing AI systems that natively manage ongoing user interaction, including real-time thinking, response, and action.
Users and prospective developers can evaluate the lab on collaboration quality and responsiveness, not solely on static model outputs or benchmark-style capability.
Second-order effects
The preview raises the bar for other model providers pursuing assistant and agent experiences: interface-layer orchestration alone may be less differentiating if interaction behavior becomes a model-level capability.
Thinking Machines can potentially pair its broad, open-weight Inkling model work with its interaction research, giving adopters both model access and a design direction for more controllable human–AI workflows.
Third-order effects
If native interaction capabilities prove useful, competition may shift from supplying standalone models toward supplying systems designed for persistent, steerable collaboration—where user judgment remains part of the operating loop.
The lab’s emphasis on shapeability suggests an emerging split in AI development between broadly capable base models and the mechanisms through which organizations adapt, supervise, and direct them.
The trend: This is one data point in the move from turn-based chatbots toward AI systems built for continuous, user-directed collaboration.
lowkey the funniest videos of the batch. thinky has some comedians!! congrats to @thinkymachines on reviving the omnimodel dream that others could not [image]
Very cool announcement from Thinky! The model looks nice (they go into some reasonable amount of detail), and reading some parts of the blog you can definitely see that the infea guys had a lot of fun there! [image]
AI changes us. Thinking Machines exists to build AI tools that increase human participation, preserve dignity across different minds, and move fast without severing society from our slower layers of memory, culture, and care. Interaction models are such a tool: an experiment in
The demo of this model is cool. Interacts with multiple people in multiple languages. On multiple tasks. Now we know why @miramurati got the big bucks.
Sharing our work on full-duplex multimodal models — real-time interaction that's natural and intuitive without compromising on intelligence. We started Thinky in part to differentially advance capabilities for human-AI collaboration, which are underemphasized relative to
Thinky's secret plan: 1: Increase Human<->AI bandwidth 2: Raise ceiling of human+AI intelligence 3: Help humans continue as main-characters in the new world We are at Step 1. Interaction Models are great real-time collaborative tools for humans. Here's a preview:
We started Thinking Machines to advance human-AI collaboration, and this is our first bet on what that looks like. Most labs treat autonomy as the goal and interactivity as scaffolding around a turn-based core. We think the way we work with AI matters as much as how smart it is.
The current “AI experience” often feels like a conversation that only begins after we stop talking. We have to batch our thoughts. We can't point at things. We phrase questions like emails. The interface doesn't leave room for us so we adapt to the models.
Today we're sharing our work on interaction models. A new class of model trained from scratch to handle real-time interaction natively, instead of gluing it onto a turn-based one. https://www.youtube.com/...
While Lilian is telling a story, the interaction model can track when she is thinking, yielding, self-correcting, or inviting a response; there is no specific built dialogue management system. [video]
People talk, listen, watch, think, and collaborate at the same time, in real time. We've designed an AI that works with people the same way. We share our approach, early results, and a quick look at our model in action. https://thinkingmachines.ai/ ... [video]