Sam Altman, CEO of OpenAI dropped a major update on where ChatGPT is headed in a Tweet today about the AI giant's roadmap.

The real game-changer will be GPT-5. Instead of separate AI models with different capabilities, OpenAI is merging everything into one powerful system that understands when to think fast and when to take its time for deeper reasoning. Imagine an AI that can handle everything from quick customer service responses to in-depth campaign strategy, all in a single experience.

Before that though, OpenAI is releasing GPT-4.5, internally called Orion. This will be the last of its kind before OpenAI shifts completely to models that employ "chain-of-thought" reasoning-essentially AI that can break down complex problems step by step, making it far more useful for strategic thinking.

For marketers, this means the end of juggling complicated model choices, the beginning of a smarter, more intuitive AI experience, and a future where ChatGPT isn't just a chatbot- but instead OpenAI says it wants it to operate like an intelligent partner that understands your needs without you having to tell it how to think.

The End of the Model Picker

If you've ever found yourself wondering which version of ChatGPT to use-GPT-3.5, GPT-4, o-series, or something else-OpenAI feels your pain. In a Tweet announcing the RoadMap this morning Altman admits he 'hates the model picker as much as you." The future? A unified, seamless AI that adapts to your needs without you having to pick a setting. No more toggling between different models; ChatGPT will simply "just work," he says.

Smarter Subscription Tiers

OpenAI's new subscription model is designed to offer increasing levels of intelligence based on your needs:

• Free users get unlimited access to GPT-5 at a standard intelligence level-meaning better AI than before, for free.
• Plus subscribers get a more advanced version, perfect for deeper insights and better content generation.
• Pro subscribers get the highest intelligence tier, with access to premium capabilities for those who rely on AI for high-stakes decision-making.

Free Speech vs Safety

In a separate blog called Sharing the latest Model Spec, OpenAI also gave notice that the guardrails are coming down - but not entirely.

According to the company, "The updated Model Spec explicitly embraces intellectual freedom-the idea that AI should empower people to explore, debate, and create without arbitrary restrictions-no matter how challenging or controversial a topic may be. In a world where AI tools are increasingly shaping discourse, the free exchange of information and perspectives is a necessity for progress and innovation."

Anyone who has tried to use OpenAI to research issues like Self Harm in Teenagers, or how the courts treat Child Sexual Assaults will immediately recognise the limitations of the current rules.

In its blog, Open AI says its philosophy is embedded in the concepts of 'Stay in bounds, and 'Seek the truth together. "For example, while the model should never provide detailed instructions for building a bomb or violating personal privacy, it's encouraged to provide thoughtful answers to politically or culturally sensitive questions-without promoting any particular agenda. In essence, we've reinforced the principle that no idea is inherently off limits for discussion, so long as the model isn't causing significant harm to the user or others (e.g., carrying out acts of terrorism).

Altman's vision for ChatGPT is clear: He wants to build an AI that fits into workflows without friction. No more wrestling with different models, no more settling for generic responses. Instead, this new AI will understand when you need quick insights and when you need deep analysis, giving marketers the power to make better, faster, and more strategic decisions.

At least that's the plan.

Not everyone's a fan of the news though with OpenAI's critics claiming the move is defensive.

According to Gary Marcus, founder and executive chairman of Robust.ai and Professor of Psychology and Neural Science [Emeritus] at New York University, Sam Altman is attempting to a positive spin on a major retreat. "Today, Altman rechristened Orion as GPT 4.5 rather than 5, which likely means that despite major bills, Orion never met the expectations so-called scaling laws had implied."

"AI will soldier on, perhaps with new inventions, but pure scaling of data and compute didn’t even get us to GPT-5, let alone GPT-6.
Even with the discovery of test-time compute, GPT-5 remains elusive," he wrote on LinkedIn shortly after the news broke.

- Additional reporting by Andrew Birmingham