OpenAI o3-mini

Pivot 5: 5 stories. 5 minutes a day. 5 days a week.

1. OpenAI o3-mini

OpenAI has released OpenAI o3-mini, a cost-efficient model that offers exceptional STEM capabilities in science, math, and coding. It supports functions calling, structured outputs, and developer messages, and offers three reasoning effort options: low, medium, and high. o3-mini is available in ChatGPT Plus, Team, and Pro users, with Enterprise access coming in February.

It is optimized for technical domains requiring precision and speed, with medium reasoning effort matching OpenAI o1's performance in math, coding, and science.

Read the full story here

2. U.S. Navy bans use of DeepSeek due to ‘security and ethical concerns’

The US Navy has issued a warning to its members to avoid using artificial intelligence technology from China's DeepSeek due to potential security and ethical concerns. The warning was issued via email to shipmates, referring to the Department of the Navy's Chief Information Officer's generative AI policy. DeepSeek's release of its R1 reasoning AI model, which rivals OpenAI's ChatGPT, has sparked concerns about future AI products requiring less-expensive infrastructure than Wall Street has assumed.

The DeepSeek model has surged to the top of Apple's App Store, dethroning OpenAI's ChatGPT. Shares of AI chipmakers Nvidia and Broadcom dropped 17% on Monday, wiped out a combined $800 billion in market cap.

Read the full story here

3. Microsoft makes OpenAI’s o1 reasoning model free for all Copilot users

Microsoft is introducing OpenAI's o1 reasoning model, Think Deeper, to all Copilot users for free. This feature allows Copilot to handle complex questions, providing step-by-step answers. Launched in October, Think Deeper is available in Copilot Labs, allowing Pro subscribers to experiment with new features.

Microsoft AI CEO Mustafa Suleyman announced the new offer, expressing excitement about the opportunity for millions of users.

Read the full story here

4. Palona that lets companies infuse their agents with unique personalities that can adjust on-the-fly to customers’ needs

Palona's AI Agents are a revolutionary solution for businesses looking to enhance their sales and customer experience. These agents learn about each customer, adapt to their needs, and sell with precision and joy. With 24/7 availability, unlimited capacity, and infinite patience, businesses can quickly create customized AI agents that boost conversion and inspire loyalty.

Palona Agents can be used in various ways, such as texting, voice, web widgets, IOS, Android, and social media. Users can also see real users' first experiences with Palona AI, with some describing their agents as authentic, kind, and helpful, even handling sensitive topics. Overall, Palona AI offers a more personalized and effective solution for businesses looking to enhance their sales and customer experience.

Read the full story here

5. AI systems with ‘unacceptable risk’ are now banned in the EU

The European Union's regulators can ban the use of AI systems they deem to pose "unacceptable risk" or harm. The EU's AI Act, approved last March, covers various use cases where AI might interact with individuals, from consumer applications to physical environments. The Act has four risk levels: minimal risk, limited risk, high risk, and unacceptable risk applications.

Companies found using these AI applications in the EU will be subject to fines, potentially up to €35 million or 7% of their annual revenue from the prior fiscal year. The fines won't kick in until August, when the competent authorities will be identified and the fines and enforcement provisions will take effect.

Read the full story here