> ## Content Index
> Fetch the complete content index at: https://nordisk-ai.ghost.io/llms.txt
> Use this file to discover other available public pages before exploring further.

# Why OpenAI's New Model Is a BIG DEAL
- URL: https://nordisk-ai.ghost.io/why-openai-s-new-model-is-a-big-deal/
- Published: 2024-09-23T17:57:06.000Z
- Updated: 2026-08-27T17:21:25.000Z
- Description: OpenAI's new model "o1" represents a major advance in AI technology through its ability to solve complex reasoning tasks. With the potential to revolutionize research and development, o1 is a step toward making AI an indispensable tool across several scientific fields.
- Author: Anders Hagen
- Tags: AI Models, AI, AI Model, ChatGPT, o1, OpenAI, Sweden, Sweden, #Migrated-1787841634512, #wp, #wp-post, #Import 2026-08-27 14:41, #Import 2026-09-07 13:41

OpenAI has launched a new model called "o1" that significantly improves the model's ability to handle advanced reasoning tasks. Unlike earlier language models, which mainly focused on text generation and language understanding, o1 is designed to tackle more complex tasks like math, programming, and scientific problems. This is a major step forward for AI technology and could potentially change how we use AI in research and development.

#### What Makes the o1 Model Unique?

The new model uses a technique OpenAI calls, and which is recognized in prompt engineering as, "chain of thought." This means the model learns to recognize and correct its own mistakes, break complex problems down into smaller steps, and try different approaches when one doesn't work. The model has performed exceptionally well in tests, such as ranking in the 89th percentile in competitive programming and among the top 500 in the USA Math Olympiad.

#### Practical Applications

The model's strength lies in its ability to solve advanced questions in fields like astrophysics and organic chemistry with greater accuracy than previous models. This could make the o1 model an invaluable resource for researchers and engineers, signaling that AI could soon become a more practical tool for problem-solving in these fields. For example, the new model has proven to be 83.3% accurate on solutions to Math Olympiad questions, compared to 13.4% for its predecessor, GPT-4o.

#### Critical Voices and Limitations

Despite the model's impressive performance, there are skeptics. Yves-Alexandre de Montjoye, professor of mathematics and computer science at Imperial College London, points out that comparisons between LLMs and human abilities aren't always meaningful, since AI models may rely on built-in knowledge rather than logically reasoning their way to an answer. Even though o1 is good at structured problems, it can still struggle with open-ended reasoning tasks, according to Google researcher François Chollet.

#### Economic Aspects and Future Development

Another challenge is cost. At $15 per 1 million input tokens through OpenAI's API, o1 is three times more expensive than GPT-4o, which could limit its use. Still, this shows the race toward AI models that can reason better than humans is underway, and we'll see its true potential once researchers and developers start exploring the model's limits.