# OpenAI Introduces GPT-4.5: Advancing AI Progress
Welcome to **[Neural](https://9to5mac.com/guides/neural/)**, your source for the latest in the dynamic AI realm. OpenAI has recently unveiled GPT-4.5, the newest version of its ChatGPT model, featuring enhanced accuracy and emotional responsiveness. Nonetheless, OpenAI has made it clear that although GPT-4.5 represents a notable advancement, it is not the pinnacle of their models thus far.
## GPT-4.5: More Intelligent and Conversational
OpenAI recognizes the uncertainty surrounding its multiple models. GPT-4.5 is presented as a research preview instead of a leading-edge model, with the robust o3 reasoning model still regarded as the most advanced. This iteration is intended to succeed GPT-4o, delivering improved accuracy, fewer hallucinations, and heightened emotional intelligence.
A highlight of GPT-4.5 is its capacity to interact with users in a more conversational style. Unlike earlier versions, this model fosters ongoing discussions rather than just dispensing information. This transition towards a clearer and more engaging interaction format meets user demands for a more instinctive conversational experience.
Starting today, GPT-4.5 is accessible to Pro subscribers for a monthly cost of $200, with plans to grant access to ChatGPT Plus ($20/month) and Team subscribers next week, followed by Enterprise and Edu users soon after. For those eager for a more in-depth look at the model, OpenAI has made available a detailed 30-page system card outlining GPT-4.5’s features.
## The Future of AI: Beyond OpenAI
As OpenAI forges ahead in innovation, former prominent figures from the organization are establishing their own enterprises, contributing to the swiftly changing AI landscape.
### Thinking Machines Lab
Mira Murati, the former Chief Technology Officer at OpenAI, has launched **Thinking Machines Lab**, aiming to develop multimodal AI models that are both practical and accessible. With a team of over 30 specialists from OpenAI, Meta, and Google DeepMind, the startup focuses on narrowing the gaps in AI understanding and personalization. Their goal emphasizes human-AI collaboration, aiming to create flexible AI systems that serve a wider array of applications.
### Safe Superintelligence
Another significant initiative is **Safe Superintelligence**, co-founded by Ilya Sutskever, the former chief scientist at OpenAI. This startup is committed to tackling the technical issues related to ensuring safe superintelligence. Their strategy merges safety and capability enhancements, maintaining a focus on safety as they advance AI development.
### Eureka Labs
**Eureka Labs**, established by Andrej Karpathy, a previous research scientist at OpenAI, is concentrating on building an AI-centric educational platform. Their initial offering, **LLM101n**, seeks to assist students in training their AI, utilizing generative AI’s features to enrich the learning process. Karpathy envisions a future where AI greatly enhances educational access and efficacy.
## Meta’s AI Aspirations
Alongside these startups, Meta is reportedly in the process of developing a standalone AI application, demonstrating its aim to become a key player in the AI field. This initiative resonates with CEO Mark Zuckerberg’s vision for advancing Meta’s AI capabilities beyond its current platforms.
## Conclusion
The introduction of GPT-4.5 signifies an exhilarating moment in the AI sector, highlighting OpenAI’s dedication to enriching user engagement through better conversational skills and emotional intelligence. As former OpenAI executives embark on fresh endeavors, competition within the AI industry is poised to escalate, promising innovative solutions that could transform our interactions with technology. Stay tuned for more updates on these developments in future issues of **[Neural](https://9to5mac.com/guides/neural/)**.
