In October, the MIT described a new approach to training artificial intelligence (AI) algorithms: “Less than one-shot,” or LO-shot learning – which could, in theory, bring a fundamental change to machine learning (ML). An AI model trained using this technique could be able to recognize even more real objects than was presented in the training dataset. In other words, training neural networks would not require such a large sample of data as was usually needed.
ICT.Moscow has interviewed Russian AI and ML experts to discuss what LO-shit learning is about and how revolutionary this approach really is.
