OpenAI launches the o1 model, a step towards human-like artificial intelligence
OpenAI launched the o1 model, which represents an advancement towards human-like artificial intelligence. The model is able to solve complex problems and improve code writing, although it is more expensive and slower than GPT-4o
2' min read
2' min read
The signal was given Sam Altman with a sibilant tweet: 'no more patience, jimmy'. And close on its heels was the official launch of 'OpenAI o1', a new series of artificial intelligence models with reasoning capabilities, released to 'solve increasingly complicated problems', the company wrote.
The new 'software' is capable of providing 'better problem solving than previous versions of the technology. This represents an early preview of this series, which includes o1-preview and o1-mini, on ChatGPT and in our APIs,' the San Francisco-based company wrote in the note.
For OpenAI 01 represents a step towards its broader goal of human-like artificial intelligence. In practice, 01 does a better job of writing code and solving complex problems than previous models. But it is also apparently more expensive and slower to use than GPT-4o .
According to the official note, through the training process 'these models learn to refine the processing method by using various possibilities and recognising errors. In our tests, these models behave like PhD students grappling with complex questions in physics, chemistry and biology. In addition, we found that the new models perform well in mathematics and coding.
Sam Altman's company also explained what are termed the 'limitations' of these models: "Being a preliminary model," OpenAI 01 "does not yet have many of the features that make ChatGPT useful, such as web browsing for information or uploading images and files. Also absent are relevant features in the API such as tool support, call function, streaming and message customisation. For many more common cases, GPT-4o remains more functional for now.
