Since then DeepSik burst onto the scene in January, Chinese open source has gained momentum artificial intelligence models. Some researchers are pushing for an even more open approach to AI, one that would allow model creation to be distributed globally.
Master IntelligenceThe startup, which specializes in decentralized artificial intelligence, is currently training an advanced large language model called INTELLECT-3, using a new kind of distributed reinforcement learning for fine-tuning. The model will demonstrate a new way to build competitive, open AI models using a variety of hardware in different locations, without the help of big tech companies, says Vincent Weisser, the company's CEO.
Weisser says the world of artificial intelligence is currently divided between those who rely on closed American models and those who use open Chinese offerings. The technology Prime Intellect is developing democratizes AI, allowing more people to create and modify advanced AI for themselves.
Improving AI models is no longer a matter of simply increasing the amount of training data and computation. Today's edge models use reinforcement learning to improve upon completion of the pre-training process. Do you want your model to excel at math, answer legal questions, or play Sudoku? Let him improve himself by practicing in an environment where you can measure success and failure.
“These reinforcement learning environments are now the bottleneck for really scaling capabilities,” Weisser tells me.
Prime Intellect has created a platform that allows anyone to create a reinforcement learning environment tailored to a specific task. The company brings together the best environments created by its own team and community to customize INTELLECT-3.
I tried running the Wordle puzzle-solving environment created by Prime Intellect researcher Will Brown, watching a small model solve Wordle puzzles (to be fair, she was more methodical than I was). If I were an AI researcher trying to improve a model, I would fire up multiple GPUs and practice the model over and over while the reinforcement learning algorithm changes its weights, thus turning the model into a Wordle master.
