Based on OpenAI CEO Sam Altman’s recent announcements, the company is making significant strides in advancing its AI capabilities with upcoming GPT-4.5 and GPT-5 releases.
Focus on GPT-4.5
The immediate focus is on GPT-4.5 (internally called Orion), which Altman describes as OpenAI’s “last non-chain-of-thought model.” This suggests a pivotal shift in how future models will process information and generate responses.
Unifying Model Offerings
A key priority for OpenAI is unifying their model offerings, specifically integrating the o-series and GPT-series models. The goal is to create systems that can intelligently utilize all available tools and determine appropriate processing times based on task complexity.
Plans for GPT-5
For GPT-5, OpenAI plans to integrate multiple technologies, including their o3 system. Notably, they will discontinue offering o3 as a standalone model, indicating a move toward more unified and comprehensive AI solutions.
Rollout Strategy
The rollout strategy includes tiered access levels:
- Free users will get unlimited chat access at “standard intelligence”
- Plus subscribers will access “higher intelligence” capabilities
- Pro subscribers will receive “even higher intelligence” features
While specific release dates weren’t disclosed, Altman indicated deployment would occur in “weeks / months.” This careful approach aligns with OpenAI’s commitment to responsible AI development and thorough testing.
User Experience Improvements
The company also acknowledges current user experience challenges, particularly with model selection. Altman noted they “hate the model picker as much as you do” and are working to return to what he calls “magic unified intelligence”—suggesting a more streamlined and intuitive user experience is forthcoming.
These developments represent significant progress in AI capabilities while demonstrating OpenAI’s focus on accessibility and practical application of their technology.
