# **Thinking Machines Lab: New AI Startup by Mira Murati**
## **Introduction**
Mira Murati, the former Chief Technology Officer (CTO) of OpenAI, has launched her own artificial intelligence (AI) startup, **Thinking Machines Lab**. This new enterprise is poised to compete with established AI leaders, including OpenAI’s ChatGPT, featuring a team largely made up of ex-senior OpenAI engineers. Although specific product details are limited, the startup’s mission statement indicates an emphasis on **multimodal AI systems, AI safety, and infrastructure development**.
## **Who is Mira Murati?**
Mira Murati significantly contributed to the creation of OpenAI’s **ChatGPT**, which is among the most utilized AI models globally. She briefly acted as **interim CEO** of OpenAI before parting ways with the organization. After her exit, it was reported that she aimed to secure **$100 million** for a new AI initiative. Now, with the formal introduction of Thinking Machines Lab, Murati is entering the competitive AI arena as a **CEO**.
## **The Team Behind Thinking Machines Lab**
Thinking Machines Lab has drawn in top-tier AI talent, especially from OpenAI. The startup’s workforce comprises around **30 researchers**, with **two-thirds** being former OpenAI engineers. Notable team members include:
– **Mira Murati (CEO)** – Former OpenAI CTO
– **Barret Zoph (CTO)** – Former OpenAI VP of Research
– **John Schulman (Chief Scientist)** – Co-founder of OpenAI
– **Alexander Kirillov** – Multimodal AI expert from OpenAI
– **Lilian Weng** – Former VP of Research (AI safety) at OpenAI
– **Luke Metz** – Ex-OpenAI researcher
– **Jonathan Lachman** – Formerly led special projects at OpenAI
Additionally, Thinking Machines Lab has onboarded engineers from **Google, Anthropic, and Mistral**, enhancing its AI development capabilities.
## **What is Thinking Machines Lab Building?**
Although the company has yet to disclose specific AI products, its **mission statement** sheds light on its focus areas:
### **1. Multimodal AI Systems**
Thinking Machines Lab strives to create AI that can **collaborate with humans** instead of operating as entirely autonomous systems. The startup champions **multimodal AI**, which merges various input types like **text, images, and audio** for more seamless and effective interactions.
> *“Instead of concentrating solely on developing fully autonomous AI systems, we are eager to create multimodal systems that collaborate with people.”* – Thinking Machines Lab
### **2. Expanding AI Applications Beyond Programming**
While many current AI models excel at **programming and mathematics**, Thinking Machines Lab aims to create AI capable of **adapting to broader domains of human expertise**. This indicates the company’s interest in **widening AI’s influence** across different sectors.
> *“We recognize the vast potential for AI to assist across all professional fields. Although current systems are proficient in programming and mathematics, we are developing AI that can adjust to the entire spectrum of human expertise.”*
### **3. AI Safety and Ethical Development**
AI safety is a pivotal issue in the industry, and Thinking Machines Lab is adopting a **preventive approach**. The company plans to enforce **rigorous safety protocols**, disseminate best practices within the industry, and support external research on AI alignment.
> *“The most effective safety measures stem from a blend of proactive research and meticulous real-world trials.”*
The company’s AI safety strategy encompasses:
– **Mitigating misuse** of AI models while preserving user autonomy
– **Sharing best practices** for AI safety throughout the industry
– **Promoting external research** by granting access to datasets, code, and model specifications
### **4. Advancing AI Infrastructure**
Thinking Machines Lab is focusing on **AI infrastructure** to guarantee **efficiency, dependability, and security**. The goal is to create **long-term solutions** instead of opting for quick fixes in AI development.
> *“Our objective is to build correctly for the long term, ensuring both productivity and security without taking shortcuts.”*
## **How Will Thinking Machines Lab Compete with OpenAI?**
Thinking Machines Lab is entering a fiercely competitive AI landscape led by companies such as **OpenAI, Google DeepMind, Anthropic, and Mistral**. Nevertheless, the startup’s **distinctive approach to AI development** could differentiate it:
– **Emphasis on Human-AI Collaboration** – In contrast to entirely autonomous AI models, Thinking Machines Lab aims to develop AI that works **in partnership with humans**.
– **Multimodal Capabilities** – The startup prioritizes **advanced multimodal AI**, potentially resulting in **more intuitive and flexible AI applications**.
– **Leadership in AI Safety** –
