Chapter 373: Car Overturning Press Conference (1) (Ch 373)

This time, the O1 model took more than a minute to display a complete, structured analysis report on the screen.

The report includes the following sections:

Introduction: The background of global climate change and the definition of extreme weather events are briefly introduced.

Data and methods: Describes the data sources (such as meteorological observation data, climate model simulation results, etc.) and analysis methods (such as statistical analysis, attribution research, etc.) used for analysis.

Changing Trends of Extreme Weather Events: Analyzes the changing trends in the frequency and intensity of global extreme heat waves, droughts, floods, storms, and other events in recent years, and cites specific statistics and charts.

The link between climate change and extreme weather events: A detailed discussion of how climate change affects the mechanisms of occurrence of different types of extreme weather events, such as:

Extreme heat: Global warming is causing average temperatures to rise, making extreme heat events more frequent and intense.

Drought: Climate change is altering the global water cycle, leading to less precipitation and more evaporation in some regions, which intensifies droughts.

Flooding: Climate change is leading to more extreme precipitation events, while rising sea levels are increasing the risk of flooding in coastal areas.

Storms: Climate change may alter atmospheric circulation patterns, affecting the paths and intensity of storms.

…

“As you can see, the third example is a complex, interdisciplinary problem that requires the integration of multidisciplinary knowledge such as meteorology, climatology, and statistics. The O1 model was able to generate a detailed analysis report based on existing scientific research and data, which demonstrates its powerful information integration, analysis, and reasoning capabilities. It can not only understand the problem, but also conduct in-depth analysis from multiple angles and draw organized conclusions.”

In the live broadcast room of platform X, comments from netizens are constantly being refreshed.

“There’s something to this answer!”

“Wow! That’s amazing!”

“OpenAI did not disappoint us!”

“This is real artificial intelligence!”

“The O1 model will change the world!”

These exclamations came like a tide, pushing the atmosphere in the live broadcast room to a climax.

But at the same time, some different voices have emerged. Some netizens compared the O1 model with Panshi Technology’s Xiaotuanzi AI model and believed that the O1 model was nothing special and even a little backward.

“It feels just so-so. Compared with Xiaotuanzi, it doesn’t seem to have any advantages,” a netizen calmly commented. He obviously has some understanding of the two AI models and made an objective comparison.

“These examples can be easily solved by little Dumpling.”

“Is OpenAI a bit behind? It’s been so long since the little dumpling was released.”

“It feels like OpenAI is just rehashing old ideas. There’s nothing new about them.”

…

“I know that you may think that the three examples just now are too simple and not enough to show the full strength of the O1 model. So, next, I will show you the performance of the O1 model on multiple mathematical benchmark tests.” The speaker is Sam Altman.

After a pause, Altman said confidently: “In these tests, the O1 model beat all currently known AI models, including Google’s Gemini series and some other open source models. The O1 model has achieved a significant lead in these tests.”

On the big screen, charts are constantly switching, showing the scores of the O1 model on different math benchmark tests, including GSM8K, MATH, MMLU, etc., which are all recognized and challenging math reasoning ability tests in the industry.