The Allen Institute for Artificial Intelligence (AI2) has released an open-source large language model OLMoE, aimed at providing high performance and low-cost solutions. This model employs a Sparse Mixture of Experts (MoE) architecture with 7 billion parameters, but through an intelligent routing mechanism, only 1 billion parameters are used per input token, achieving efficient computation. OLMoE includes both general and instruction-tuned versions, supporting a context window of 4096 tokens. Its training data is sourced from a wide range, including Common Crawl, Dolma CC, and Wikipedia.

At the 2024 World Robot Conference, Zhi Ping Fang (Shenzhen) Technology Co., Ltd. launched its new universal intelligent robot, Alpha Bot 1S. This robot is equipped with Zhi Ping Fang's self-developed universal embodied intelligent system, AI2R Brain, providing high flexibility, perception capabilities, and operational range. Through an innovative combination of wheeled design, folding lift structure, and high-degree-of-freedom robotic arms, Alpha Bot 1S shows significant improvements over its predecessor in terms of joint count, sensor count, and battery life. Alpha Bot 1S acts as an AI2

Recent turmoil at Meta: R-CNN author Ross Girshick has decided to leave and join AI2. There has been a wave of departures from the FAIR team, including the author of ResNeXt who has returned to academia. Meta is facing challenges in retaining AI talent, even though there have been advancements in open source. Yann LeCun, Meta's Chief Scientist, has expressed understanding regarding the departures, emphasizing that FAIR encourages scientists to publish papers. Girshick's departure from Meta FAIR may impact its...

AI21 has announced the completion of its Series C funding, raising $53 million, with a total valuation of $1.4 billion and a total funding amount of $208 million. Amid internal turmoil at OpenAI, AI21 has chosen to announce its funding to compete for market share in generative AI. AI21 emphasizes close collaboration with enterprises, focusing on customization, robustness, and security, offering tailored solutions that differ from OpenAI. AI21 has performed well in the competition, welcoming investments from Intel Capital and Comcast Ventures.

Israeli startup AI21 Labs announced the completion of $155 million in funding, with Google and Nvidia also participating. AI21 Labs focuses on customized generative AI models, boasting its own large model, Jurassic-2, competing with OpenAI to provide more accurate and reliable text generation services.

The co-founder of AI21 Labs stated that they usually win in the competition with OpenAI for enterprise AI business. OpenAI's tools, such as ChatGPT, often do not perform stably enough to meet enterprise demands. AI21 Labs focuses on improving model stability and interpretability while providing custom language models.

AI21 Labs has completed a $155 million Series C funding round, reaching a valuation of $1.4 billion. AI21's proprietary Jurassic-2 model is considered one of the largest LLMs in the world. Recently, AI21 has collaborated with clients across various industries and has been selected for the inaugural CB Insights GenAI50 list.

AI2 recently released an open source dataset named Dolma, which contains 3 trillion tokens. Dolma's data will serve as the foundation for AI2's developing Open Language Model OLMo, expected to launch in early 2024. The Dolma dataset comes from a wide range of sources, including web content, academic publications, code, and books, making it the largest publicly available dataset of its kind.

["The Allen Institute for Artificial Intelligence has released the open text dataset Dolma, aimed at promoting transparency and innovation in AI language models.", "Dolma, as the core of AI2's open language model initiative, provides researchers and developers with free data resources.", "The Dolma dataset boasts a scale of 3 billion tokens, operates under the mid-risk ImpACT license, and encourages users to provide contact information and usage feedback."]