Thinking Machines Launches Inkling AI Model

Thinking Machines, an AI lab established last year by former OpenAI leader Mira Murati, has announced the release of its inaugural AI model, Inkling. This launch positions the startup to compete with companies like Anthropic by focusing on an "intellectual vibe" for AI development. Inkling, which is an open-weights model, features 975 billion parameters, with 41 billion actively utilized. The model can be further customized through the company's Tinker platform, which was made available last fall.

While Thinking Machines acknowledges that Inkling is not the most powerful AI model currently available, the company's strategy centers on offering high levels of customization. This approach aims to differentiate itself in the market by allowing consumers and enterprises to tailor the AI to their specific needs and preferences, rather than relying on a system that reflects generalized user data. The company's stated goal is to enhance human intelligence rather than replace it, a message that echoes some of the branding and philosophy espoused by Anthropic.

Anthropic has consistently positioned itself as an "intellectual" and "creative" AI platform, utilizing advertising and influencer partnerships to convey this message. Their AI models are named after literary forms, and CEO Dario Amodei has cultivated a thought leader image through his blog posts. The company's core message is that AI should augment human cognition. Thinking Machines' emphasis on "cultivating" unique organizational knowledge through AI, as stated in a recent company blog, aligns with this vision of AI as a collaborative tool rather than a replacement for human expertise.

The company's blog post articulated this philosophy, stating, "For artificial intelligence to benefit from distributed knowledge, it must itself be distributed. Every organization is powered by the expert knowledge of its people, gained and expressed through their work." Thinking Machines believes in AI that helps organizations develop their unique knowledge base, contrasting this with AI that might simply extract and replicate a standardized offering. This approach underscores their commitment to AI that works alongside people, supporting an ongoing process of knowledge cultivation.