AI Battle: Rozetked Pits AI Assistants Against Each Other
Rozetked moved beyond boring benchmark numbers and challenged AI assistants to solve everyday human tasks. The author tested Alice, GigaChat, and ChatGPT on erudition, creativity, code, computational ability, and empathy.
Our Alice only slightly lagged behind the Western leader: it wrote working code and pleased with a user-friendly interface. And if we take into account that ChatGPT is unavailable to most Russian users, then it generally showed itself as a top option.
Read more about how the comparison went AI-турнир на Rozetked.
AI Battle: Rozetked Pits AI Assistants Against Each Other
Rozetked moved beyond boring benchmark numbers and challenged AI assistants to solve everyday human tasks. The author tested Alice, GigaChat, and ChatGPT on erudition, creativity, code, computational ability, and empathy. This bold experiment showed how different AI models cope with real-world challenges.
Our Alice only slightly lagged behind the Western leader: it wrote working code and pleased with a user-friendly interface. And if we take into account that ChatGPT is unavailable to most Russian users, then it generally showed itself as a top option. GigaChat, in turn, demonstrated impressive results in generating texts and answering questions, although it was inferior in some aspects.
The comparison revealed the strengths and weaknesses of each model. For example, ChatGPT led in creative tasks, but Alice proved to be more understandable and faster in everyday use. GigaChat has become an excellent choice for those who need Russian-language support and integration with services.
During the testing, the following were evaluated:
- Erudition: the ability of AI to answer complex questions.
- Creativity: generation of texts, poems, scripts.
- Code: writing and debugging program code.
- Calculations: solving mathematical problems.
- Empathy: understanding and responding to emotional context.
Read more about how the comparison went AI-турнир на Rozetked. Discover More!