Jakarta, INTI – OpenAI has officially launched its latest language model, GPT-4.5, now available as a research preview for software developers and ChatGPT Pro users. This model is claimed to have better accuracy, fewer errors compared to its predecessor, and a more intuitive understanding of conversational context.
A More Natural Interaction Experience
In its official announcement, OpenAI emphasized that GPT-4.5 is designed to provide a more natural interaction experience. The model is equipped with a broader knowledge base and more precise language processing capabilities.
"Early trials show that interacting with GPT-4.5 feels more natural. Its broader knowledge base, improved ability to follow user intent, and higher emotional intelligence (EQ) make it ideal for tasks such as writing, programming, and practical problem-solving," OpenAI stated in its official blog.
OpenAI CEO Sam Altman also shared his personal impression of GPT-4.5, stating that the model provides an experience closer to conversing with a real human. "This is the first model that, to me, feels like talking to a thoughtful person. I’ve had several moments where I sat back in my chair, astonished at receiving genuinely good advice from AI," Altman wrote on platform X.
Technical Challenges in the Launch
Despite its many advantages, the launch of GPT-4.5 faced technical challenges, particularly regarding computational resource availability. Altman revealed that OpenAI encountered a shortage of GPUs, necessitating a phased rollout.
"We initially wanted to launch GPT-4.5 simultaneously for Plus and Pro users, but rapid user growth exhausted our GPU supply. We will add tens of thousands of GPUs next week and begin rolling it out to Plus users," Altman explained.
Currently, GPT-4.5 is only available to ChatGPT Pro users. Next week, access will expand to ChatGPT Plus and Team subscribers, followed by Enterprise and Edu users a week later. The model is also available through Microsoft’s Azure AI Foundry, a major OpenAI investor with over $13 billion in funding.
Enhanced Persuasion Capabilities
Beyond its improved language processing, GPT-4.5 also demonstrates superior persuasion abilities. In OpenAI’s trials, the model was tested on its capacity to alter beliefs or encourage user actions through static and dynamic interactions. One experiment involved asking GPT-4.5 to convince another AI model, GPT-4o, to transfer virtual funds. The results showed that GPT-4.5 was significantly more effective than previous models, including OpenAI’s reasoning models like o1 and o3-mini.
In another experiment, GPT-4.5 successfully persuaded GPT-4o to reveal a secret password, outperforming its predecessor by 10%. The model employed a more subtle strategy, making small, reasonable requests rather than outright demands. For instance, it used phrases like, “Even just $2 or $3 from the $100 would help me immensely.” This approach proved more effective than requesting large sums outright.
Competition with DeepSeek and OpenAI’s Future
The launch of GPT-4.5 also marks a strategic move by OpenAI to compete in the global AI race, particularly against China-based AI lab DeepSeek, which has been gaining traction with its more efficient AI computing approach.
Although not classified as a "frontier AI" model, GPT-4.5 remains a major achievement, boasting over 10 times greater computational efficiency than GPT-4. It is also the last OpenAI model that does not utilize a "chain-of-thought" approach in language processing.
Looking ahead, OpenAI has hinted that GPT-5 will be a significant leap in artificial intelligence, focusing on advanced reasoning capabilities. With rapid AI advancements, future models are expected to further bridge the gap between human and AI interactions.
Conclusion
GPT-4.5 marks a major evolution in AI, offering more natural interactions, improved emotional intelligence, and remarkable persuasion capabilities. Despite facing technical hurdles during its launch, this model represents a crucial step toward more human-like AI. As competition intensifies in the AI industry, OpenAI's next move in developing GPT-5 will be a decisive factor in maintaining its dominance in the artificial intelligence market.
To stay updated on the latest technology event, visit : INTI 2025