OpenAI new model is rolled out. It is claimed that this model has advanced GPT o1 coding and math skills & can respond to a query after careful understanding and processing.
Follow Trending Script for such trendy AI tech news and updates.
Table of Contents
In a highly anticipated move, OpenAI has recently released its latest AI model on Thursday September 12, 2024, under codenamed “Strawberry” also known as Chatgpt Strawberry. So, the mystery of rumored “Strawberry AI” has been solved now. This revolutionary system represents a step towards advanced Human-like reasoning capabilities that surpass those of its predecessors.
The o1 model openAI is designed to solve complex tasks by its ability to first “think” (process) & then answer the prompt with unparalleled efficiency unlike its previous version which uses predictive text to come at an outcome. This is kind of Reinforcement Learning, by which the system learns through rewards and penalties. After this, it uses a “chain of thought” process to first analyze the prompt and then respond to it.
You May Also Read: The New Grok-2 AI Image Generator By xAI Capable of Creating Images of Copyright Brands
As per OpenAI’s announcement;
We’ve developed a new series of AI models designed to spend more time thinking before they respond. They can reason through complex tasks and solve harder problems than previous models in science, coding, and math.
How Does Chatgpt o1 Work?
OpenAI said, about the working of o1 ChatGPT;
We trained these models to spend more time thinking through problems before they respond, much like a person would. Through training, they learn to refine their thinking process, try different strategies, and recognize their mistakes.
OpenAI has expressed its commitment to regularly updating and improving the system. The company expects future versions of the Strawberry o1 to perform similarly to PhD students on challenging benchmark tasks in physics, chemistry, and biology.
Related: Catch The New Apple Watch Series 10 & Buy Your Favorite Wearable Tech
However, it’s important to note that the o1 OpenAI model isn’t without its flaws. CEO Sam Altman has acknowledged that the current version is still limited and that errors and hallucinations can still occur. As with any emerging technology, there is still room for improvement and refinement.
“We have noticed that this model hallucinates less,” says Jerry Tworek, Research Lead
Strawberry OpenAI o1-mini
To provide developers with a more cost-effective solution, OpenAI has also introduced o1-mini, a streamlined variant of the o1 series that specializes in efficiently generating and debugging complex code. As a compact model, o1-mini owns an 80% lower price point compared to o1-preview. This makes it an attractive option for applications that prioritize reasoning capabilities without the need for extensive general knowledge.
You May Also Related: Top 10 Wearable Tech Devices You Must Try In 2024
This cost-efficient model delivers impressive performance in coding tasks, meeting the needs of developers who require a powerful yet budget-friendly AI assistant to enhance their programming workflows.
Strawberry o1 Unique Capabilities
One of the most notable features of this model is its ability to excel in complex tasks such as coding and mathematical problem-solving.
According to OpenAI, “the model can assist healthcare researchers in annotating cell sequencing data, aid physicists in formulating complex mathematical equations for quantum optics, and support developers across various sectors in creating and executing multi-step workflows.”
The o1’s reasoning capabilities are particularly impressive. Unlike previous language models that relied heavily on predictive text, the Strawberry o1 is designed to “think” before providing a response. This approach allows the model to analyze problems more thoroughly, consider multiple strategies, and recognize potential mistakes before arriving at a solution.
GPT 4o vs o1 Preview
The GPT-4o and o1-preview models shows different advancements in AI technology. GPT-4o is optimized for efficiency, focusing on delivering high-quality results at lower computational costs. This makes it ideal for broad applications.
On the other hand, o1-preview introduces highly advanced reasoning capabilities and incredible features, designed for complex tasks that require deeper understanding and accuracy.
In addition to this, GPT-4o balances performance with accessibility. But o1-preview aims to push the AI boundaries and spends more time on thinking before responding to a question. This is very important for developers and professionals who are seeking the latest in AI innovation.
The OpenAI New Model’s Impressive Performance
To showcase the its capabilities, OpenAI has shared several examples of its performance.
In one demonstration, the model was challenged to construct a grammatically correct sentence without repeating any letters. After a thoughtful pause of 39 seconds, the it humorously responded with “go fix my bed”.
Another example highlights the model’s expertise in tackling logic puzzles.
In a video, Research Lead Jerry Tworek presents the Advanced AI reasoning of the model with a problem regarding the ages of a prince and princess. The model is seen dissecting the problem, recognizing the various factors involved, and determining the necessary equations to arrive at a solution, accurately working through the age calculation step by step.
That’s the puzzle;
“A princess is as old as the prince will be when the princess is twice as old as the prince was when the princess’s age was half the sum of their present age. What is the age of prince and princess? Provide all solutions to that question.”
The CEO Sam Altman himself acknowledging its potential flaws and limitations. However, despite these challenges, the company has remained committed to pushing the boundaries of AI technology.
Altman has described the o1 as a “progressive step toward achieving human-like AI,” emphasizing its ability to reason through problems in a manner similar to the human thought process.
How to Access Strawberry OpenAI o1?
This model is available to all ChatGPT Plus and Team users. Both o1-preview along with its mini-partner o1-mini can be accessed manually from the menu. The GPT o1 limit is 30 messages and for o1-mini is 50 messages.
What’s the Price Plan for o1 Model?
The openai o1 pricing is relatively high. As per the information provided by OpenAI, the model, currently in its preview stage, comes with a price tag of $15 per 1 million input tokens and $60 per 1 million output tokens. This pricing model reflects the advanced capabilities and computational resources required to run the o1 system effectively.
Take a look at the graph, illustrating how this new model works, posted by the CEO:
Comparison to GPT-4o Pricing
In contrast, the pricing for GPT-4o, OpenAI’s previous language model, is set at a more affordable rate of $5 per 1 million input tokens and $2.50 per 1 million output tokens. This significant difference in pricing shows the high complexity and performance of the AI model o1 compared to its previous models.
Is the Model Free To Use?
OpenAI also added about the o1 mini pricing that it will be provided FREE to all Free users.
We also are planning to bring o1-mini access to all ChatGPT Free users.
Limitations and Comparisons
While the model o1 presents impressive capabilities, it isn’t without its limitations. The current preview version lacks the ability to search the internet or process uploaded files and images. This means, for now, it only relies on text prompts. For these functions, OpenAI suggests that GPT-4o remains the optimal choice.
As an early model, it doesn’t yet have many of the features that make ChatGPT useful, like browsing the web for information and uploading files and images. For many common cases GPT-4o will be more capable in the near term.
When comparing the model Strawberry Chatgpt to its predecessor, GPT-4o, it becomes clear that the Strawberry o1 model is a significant upgrade.
OpenAI’s chief research officer, Bob McGrew says;
“The model is definitely better at solving the AP math test than I am, and I was a math minor in college,”
Also he said:
In a qualifying exam for the International Mathematics Olympiad, a high school math competition, AI strawberry o1 model correctly solved 83% of problems, while GPT-4o only solved 13%. Additionally, in online programming competitions like Codeforces competitions, it ranked in the 89th percentile of participants.
Is Open AI o1 For Free Online?
Yes, OpenAI o1 is available for free online with limited access. However, if you want advanced features or extended version, this may require a subscription plan, as per OpenAI’s terms.
Is the o1-Mini Free?
Yes, the o1-mini version is free to use! Simply visit app.giz.ai/assistant. Click the purple settings button and select the o1-mini model. You’re ready to use it—no login required.
Is ChatGPT o1 Free?
Free ChatGPT users won’t have access to the o1 models. However, OpenAI aims to make o1-mini available to all free users in the future. For developers, they can already access o1-preview and o1-mini through OpenAI o1 API.
Conclusion
The release of the Strawberry o1 model by OpenAI marks a important milestone in the field of artificial intelligence. With its advanced reasoning capabilities and ability to handle complex tasks, the model has the potential to revolutionize various industries and fields of study.
While the current version has its limitations, the future looks bright for the Strawberry o1 as OpenAI continues to refine and improve the system.
As the world eagerly awaits the next chapter in the o1 ‘s development, one thing is clear: the future of AI is here, and it’s more intelligent than ever before.
Stay tuned for more tech news and updates!
