As the competition for artificial intelligence intensifies, Alphabet company is joining forces with DeepMind to combine an internal Google Research team known as Brain. Google company merges its AI-focused groups, and several high-profile projects have already been completed by the teams that will be combined, including the transformer, a technology that serves as the foundation for some of OpenAI's research. The creation of "multimodal" AI, akin to OpenAI's most recent model, GPT-4, will be prioritized moving forward by the united teams within Alphabet. This kind of AI can create new material in response to inputs that include both text and images.

London-based DeepMind, a division of Alphabet and the parent company of Google are well known for consistently introducing advances from AI-focused groups. Examples include its work on AlphaFold, a system that can predict the shape of proteins, and AlphaGo, a program that has taught itself to play Go better than any human on Earth. Internally, the organization has traditionally been viewed as a team that develops ideas for artificial intelligence that might not directly apply to Google products.

In the meantime, "transformer" technology-important building blocks for sizable language models-was developed by Google Research. The latest generation of chatbots, such as Google's Bard and OpenAI Inc.'s ChatGPT, are powered by that technology.

Now that Google has reorganized, it seems to have combined all of this research under Google DeepMind, indicating a closer synergy with the rest of Alphabet. During Alphabet's fourth-quarter revenue report in February, the company announced that DeepMind would be included in Alphabet's corporate costs beginning this year to reflect how the technology is being used in other businesses rather than being included in the "Other Bets" category of investments with less immediate impact.