Great to catch up with an amazing friend and mentor today. He has been in this business for many years and I am truly grateful for all the guidance and support he has given me along the way. Thank you for always being there to help!

Wow, another frontier model release by OpenAI, another massive jump! The LLM war continues...

GPT-5.2 Thinking is here and the benchmarks are impressive:

- 100% on AIME 2025 (competition maths)
- 55.6% on SWE-Bench Pro (software engineering)
- 86.2% on ARC-AGI 1 (abstract reasoning)
- 92.4% on GPQA Diamond (science questions)

OpenAI is pushing hard on reasoning capabilities. The competition between OpenAI, Anthropic, and Google keeps heating up!