Sam Altman, CEO of OpenAI, has unveiled his company’s roadmap regarding the future of its generative AI models, highlighting the upcoming releases of both GPT-4.5 and GPT-5. In addition to this, Altman has assured that OpenAI will simplify its product offerings to provide an enhanced experience for its users.
In this regard, the company has decided to do away with its model selector menu, which allows the activation of one AI or another for each task based on their capabilities. This process can be complicated for those less familiar with the various OpenAI models and their differences and qualities.
“We want AI to “simply work” for you; we know how complicated our model and product offerings have become. We hate the model selector as much as you do and we want to return to magical unified intelligence,” Altman assured in his statement published on Twitter.
OPENAI ROADMAP UPDATE FOR GPT-4.5 and GPT-5: We want to do a better job of sharing our intended roadmap, and a much better job simplifying our product offerings. We want AI to “just work” for you; we realize how complicated our model and product offerings have gotten. We hate… — Sam Altman (@sama) February 12, 2025
OPENAI ROADMAP UPDATE FOR GPT-4.5 and GPT-5:
We want to do a better job of sharing our intended roadmap, and a much better job simplifying our product offerings.
We want AI to “just work” for you; we realize how complicated our model and product offerings have gotten.
We hate…
— Sam Altman (@sama) February 12, 2025
The first step in OpenAI’s plan is to launch GPT-4.5, a model known internally as Orion, whose presentation has been postponed, as it was initially scheduled for December 2024. The company has even stated that it could be up to 100 times more powerful than GPT-4.
Altman has stated that this is their last “non-chain-of-thought” model. This means that their future creations will be based on this technique that generates responses broken down into intermediate steps, simulating logical reasoning.
Once GPT-4.5 is available, OpenAI will aim to unify its different models, both those of the series and the GPT range. This will be done “through the creation of systems that can use all our tools, knowing when to think for a long time or not and, in general, being useful for a wide range of tasks.”
At the end of 2024, OpenAI stated that they expected to release their model o3 at the beginning of 2025, shortly after activating o3-mini. However, this has changed and the developer no longer plans to make this launch independently, but rather it will be integrated into GPT-5.
“Both in ChatGPT and our API, we will launch GPT-5 as a system that integrates much of our technology, including o3”. It is worth noting that users of the free plan will have access to GPT-5, although with a standard intelligence configuration. Moreover, the model will be subject to abuse thresholds, which are limits or restrictions established to prevent misuse or excessive use of the platform.
Finally, Altman has stated that the Plus and Pro plans will offer versions of GPT-5 with higher levels of intelligence. “Plus subscribers will be able to run GPT-5 with a higher level of intelligence, and Pro subscribers will be able to run GPT-5 with an even higher level of intelligence. These models will incorporate voice, canvas, deep search, research, and more.”
Photo: GPT4
Your email address will not be published. Required fields are marked *
Δ