Major Advancements in AI Capabilities
OpenAI recently launched a new version of its popular AI tool, ChatGPT, marking a significant milestone in the field of artificial intelligence. The latest iteration, known as ChatGPT-5, has been rolled out for free to all users, making it accessible to nearly 700 million people who use the platform weekly.
Sam Altman, co-founder and CEO of OpenAI, highlighted that this new model represents a major leap forward in AI capabilities. He described it as "clearly a model that is generally intelligent" and emphasized that it's a substantial step toward achieving truly capable AI systems. However, he also noted that there is still a long way to go before reaching the level of artificial general intelligence (AGI), which would mimic human-like thinking and learning.
Enhanced Performance and Expertise
According to Michelle Pokrass from the development team, GPT-5 excels at acting as an independent agent, handling computer tasks with greater efficiency. Altman compared the evolution of the models over time, noting that GPT-3 felt like talking to a high school student, while GPT-4 was more like engaging with a college student. Now, GPT-5 offers the experience of conversing with a PhD-level expert on any subject.
This advancement has led to expectations of a new era in software development, dubbed "vibe-coding." Altman believes that the ability to create software programs on demand will be a defining feature of this new phase. An example provided by OpenAI executives involved the bot being asked to develop an app for learning French, showcasing the model’s versatility and depth.
Competitive Landscape and Future Goals
The release of ChatGPT-5 comes amid fierce global competition in AI technology. Companies like Google and Microsoft are investing heavily in their own AI systems, aiming to stay ahead in the race. Altman expressed confidence that ChatGPT-5 leads the pack in areas such as coding, writing, and healthcare. He also mentioned that there are "orders of magnitude more gains" to come on the path toward AGI, although this will require significant investment in computing power.
Safety and Trustworthiness
OpenAI has placed a strong emphasis on ensuring the safety and reliability of ChatGPT-5. Alex Beutel, lead researcher on AI safety, explained that the model was trained to be honest and avoid providing harmful information. Evaluations were conducted to measure the prevalence of deception, and the model was adjusted to generate "safe completions."
Government and Public Access
The launch of ChatGPT-5 also marks a shift in accessibility. OpenAI announced that the US government will have access to a business-oriented version of ChatGPT for a year at a minimal cost. Federal workers in the executive branch will benefit from this partnership with the US General Services Administration, effectively receiving the service for free.
In addition, OpenAI released two new AI models—gpt-oss-120b and gpt-oss-20b—that can be downloaded and modified by users. These open-weight language models aim to challenge similar offerings from both US and Chinese competitors. The move reflects OpenAI’s commitment to transparency, especially considering its origins as a nonprofit organization.