Thinking Machines says its mission is to build AI that people and organizations can shape and make their own, and that “extends human will and judgment”
The mission of Thinking Machines is to build AI that extends human will and judgment. — Artificial intelligence …
Thinking Machines Lab
Context & Ripple Effects
Thinking Machines Lab launched under CEO Mira Murati with a leadership team drawn from OpenAI, then described “interaction models” built for continuous, real-time collaboration between users and AI. Its stated mission places that earlier product direction inside a broader design philosophy: AI should remain adaptable to users and organizations rather than operate as a fixed, one-way system.
The significance is chiefly strategic positioning. In a field often framed around model capability alone, Thinking Machines is making user and organizational control part of the intended product identity.
First-order effects
Thinking Machines gains a clearer criterion for its model and product work: emphasize systems that users and organizations can tailor and use collaboratively.
Potential customers and developers get a more explicit signal that the lab is targeting configurable, human-in-the-loop AI experiences rather than narrowly autonomous behavior.
Second-order effects
The interaction-model approach raises the importance of interface design, customization mechanisms, and collaboration workflows alongside raw model performance.
Competing AI vendors pursuing organizational adoption face added pressure to show how users can direct, adapt, and retain judgment over their systems.
Third-order effects
If this positioning is carried into products, enterprise AI competition could increasingly differentiate on controllability and fit with human workflows, not only benchmark capability.
The broader shift is toward AI systems designed as ongoing collaborators; whether that becomes a durable market distinction depends on whether such control can be delivered without sacrificing usefulness.
The trend: This is one data point in the move from standalone generative models toward configurable, continuously interactive AI designed to augment human decision-making.
We started Thinking Machines a year and a half ago with a couple of instincts: that people should have much more ability to customize models and do research on them, and that even as AI becomes more autonomous, there's a lot more to build to make humans and AIs work well
The most interesting debate in AI is not simply whether models should learn human judgment, but where that judgment should live. The state of the art today is hybrid. Broad human preferences are encoded into model weights through supervised fine tuning, RLHF, DPO, Constitutional
Today we share the worldview behind our mission. Human values don't average out. Local knowledge can't be centralized. The good future has many AIs, raised in different places, shaped by the people they serve, disagreeing with each other the way we do. https://thinkingmachines.ai…
no wonder Thinking Machines has the most immaculate vibes of any frontier lab/neolab Hayekian AI alignment: local knowledge, decentralized power, pluralism, and humans shaping the machines — not the other way around [image]
Thinking Machines exists to pursue a differentiated view of the future of AI. Specifically, we care about those branches of the tech tree that are involved with maintaining the best of human participation in the new world in which we find ourselves. It is our belief that this is
What do we do at @thinkymachines: Personalization/sovereignty, Human Participation, Decentralization. Democratize AI and make it useful for people. All three of them reduce society's dependence on centralized AGI companies (including ours when we get important), and that is a
This essay seems to assume away the eventual existence of vastly superhuman AI systems, instead inhabiting a future where we somehow humans stay necessary in most economically-useful acitivities. I found it pretty hard to engage with because it keeps oscillating between “belief-…
thinking machines is actually such a funny company the dashboard is slow enough to be unusable but you can now train a model by dropping the api key and telling fable to train one so it doesn't matter it's literally train models by api, but more flexible than what oai has had
A year ago we set out to empower humanity with a focus on multimodal AI, custom models, and open science. We previewed interaction models that collaborate the way people do. Tinker lets anyone train their own open weights models. We published our research on Connectionism.
We're building AI that people and organizations can shape and make their own. AI should extend our will and judgment instead of neglecting it; enabling that is the technical challenge we are working to solve. https://thinkingmachines.ai/ ...