This week, the AI industry saw several significant developments in product launches, updates, and funding news from artificial intelligence-focused startups and companies.
OpenAI launched its new reasoning model, o3-mini, which offers medium-level reasoning capabilities to ChatGPT’s free users. The model responds 24% faster than its predecessor and has a lower cost. External testers preferred the responses of o3-mini over its competitor by 56%, finding 39% fewer mistakes on real-world questions.
The AI startup also introduced ChatGPT Gov, a version of its chatbot designed for U.S. government agencies to access its frontier AI models. This infrastructure aims to expedite internal authorization and enhance data handling.
Additionally, OpenAI announced an agreement with the U.S. National Laboratories to share its reasoning models for scientific research. Voice AI model developer ElevenLabs secured $180 million in funding, valuing the startup at $3.3 billion, to support its AI audio technology and develop new products.
The Allen Institute for AI released its next-generation open-source model, Tülu 3 405B, which applies fully open post-training recipes to large models. This marks Ai2’s first successful application of its Reinforcement Learning from Verifiable Rewards framework to improve the model’s performance on benchmarks like MATH.
Source: https://qz.com/openai-o3-mini-ai-model-chatgpt-gov-elevenlabs-series-c-1851753079