OpenAI has recently perfected its flagship algorithm, GPT-4, and is currently in the process of developing its successor, GPT-5. Google DeepMind’s CEO Demis Hassabis has revealed that the next big model, the Gemini, is currently in development and will combine the strengths of AlphaGo-type systems with the language capabilities of large models. Recent advances in AI have been due to algorithms becoming increasingly large and consuming more data, however, the days of large-scale AI models are over as training costs and computing power have skyrocketed.