The emergence of Large Language Models (LLMs) has marked a significant milestone in the advancement of artificial intelligence and natural language processing [1, 2, 3]. These powerful models, boasting billions of parameters and trained on massive amounts of text data, have demonstrated remarkable abilities in tasks such as language generation, question answering, and reasoning, surpassing human performance in some cases [4, 5]. The rapid progress in capabilities has sparked excitement and speculation about their potential to enable more intelligent and human-like AI systems, with some researchers even suggesting that they could be the key to achieving Artificial General Intelligence (AGI) [6, 7]. Breakneck advancements in LLM capabilities have continued with the introduction of GPT-4 level models, such as Anthropic's Claude 3.5 Sonnet [8], Claude 3 Opus [9], Google Deepmind's Gemini 1.5 Pro [10], Llama 3 405b [11] and GPT4o [12]. These state-of-the-art models have demonstrated even more impressive performance across a wide range of tasks, showcasing their potential to revolutionize various industries and research domains. Very recently ChatGPT o1-preview and o1-mini were released [13], emulating a system that looks like system II thinking.