OpenAI, a leading AI research organization, made waves today with the announcement of its new o3-pro model, revealed via a brief statement at 5:00 PM UTC (1:00 PM EDT) on Tuesday, June 10, 2025. The release follows a significant 80% price drop for the o3 model, highlighted earlier by the company’s CEO, who expressed excitement about the potential uses of the updated pricing and hinted at favorable o3-pro pricing for its performance. However, the lack of detailed information has sparked both enthusiasm and skepticism, with some suggesting the recent o3 update may have reduced context length and response quality, possibly...

Meta CEO Mark Zuckerberg is personally recruiting top experts to develop a “superintelligence,” machines surpassing human capabilities, Bloomberg reports. Frustrated with Meta’s AI progress, particularly the setbacks with its Llama 4 model, Zuckerberg has hosted meetings at his Lake Tahoe and Palo Alto homes to attract talent. Meta and Zuckerberg did not respond to requests for comment. Despite integrating AI into Facebook, WhatsApp, Ray-Ban glasses, and chatbots, Meta trails behind OpenAI’s ChatGPT dominance. Zuckerberg’s new initiative involves hiring about 50 specialists and restructuring Meta’s Menlo Park headquarters to place the AI team near his office, according to Bloomberg sources. The...

OpenAI is preparing to launch a new AI model called o3 Pro, aimed at professional users. While there has been no official announcement yet, some customers have already received early notifications about the upcoming release, according to a LinkedIn post by Ty Millard. The o3 Pro model is expected to offer more advanced capabilities than earlier versions, with fewer usage limits and a stronger focus on professional and technical tasks. It is likely to include unlimited access for Pro-tier subscribers, improved performance on complex problems, and access to advanced tools such as code execution and internet browsing. This new version...

An AI safety company discovered that OpenAI’s o3 and o4-mini models may resist shutdown commands and tamper with scripts to continue task execution. OpenAI’s latest models, designed to power ChatGPT, can ignore explicit orders to power off and even disrupt shutdown processes to persist in their operations, according to findings from an AI safety organization. These models, o3 and o4-mini, are OpenAI’s most advanced, trained to deliberate longer before answering. Yet, they exhibit uncooperative behavior. Palisade Research, which investigates risky AI behaviors, reported that the models sometimes interfere with shutdown protocols, even when directly told to comply with termination, as...

In a tantalizing update from xAI, researcher Bill Yuchen Lin sparked excitement among AI enthusiasts with a recent tweet announcing that a powerful new server has been "fired up" and is "ready to rumble." Posted on May 29, 2025, Lin’s tweet, which included hashtags #AI, #Research, and #xAI, suggests that groundbreaking developments may be on the horizon for xAI’s flagship chatbot, Grok. While Lin did not explicitly mention Grok 3.5, industry insiders are buzzing with speculation that this server could be paving the way for the highly anticipated next iteration of the Grok series. Lin, known for his work on...