# Richard Socher

Richard Socher (born 1983) is a German-born computer scientist and entrepreneur known for AI research, co-founding You.com and Recursive, and developing GloVe word embeddings.

Richard Socher (born 1983) is a German-born computer scientist specializing in [artificial intelligence](https://www.wikiprompt.org/wiki/artificial-intelligence) research. He is the co-founder and CEO of You.com, an AI search infrastructure company, and co-founder and CEO of Recursive, a company pursuing recursive self-improvement. He is also a co-founder and investment partner of AIX Ventures, an AI venture capital firm. Socher is known for his contributions to [deep learning](https://www.wikiprompt.org/wiki/deep-learning), including the development of GloVe word embeddings and the founding of MetaMind, an AI startup acquired by Salesforce.

Socher was born in 1983 in Dresden, Saxony, Germany. He earned degrees from the University of Leipzig and Saarland University before completing his PhD in computer science at [Stanford University](https://www.wikiprompt.org/wiki/stanford-ai-lab) in 2014. His doctoral work focused on recursive neural networks for natural language processing, building on advances in [neural networks](https://www.wikiprompt.org/wiki/neural-network) and [machine learning](https://www.wikiprompt.org/wiki/machine-learning).

## Early Career and ImageNet

In 2009, while at Stanford, Socher collaborated with Fei-Fei Li and a team of researchers to develop ImageNet, a large visual database designed for visual object recognition research. ImageNet hand-annotates images to indicate objects pictured, with bounding boxes provided in at least one million images. The database contains more than 20,000 categories, with typical categories such as "balloon" or "strawberry" consisting of several hundred images. The annotations of third-party image URLs are freely available, though the actual images are not owned by ImageNet. This dataset became a cornerstone for training [convolutional neural networks](https://www.wikiprompt.org/wiki/convolutional-neural-network) and spurred progress in computer vision.

## GloVe and Academic Contributions

In 2014, Socher co-authored GloVe, an open-source [unsupervised learning](https://www.wikiprompt.org/wiki/unsupervised-learning) algorithm that embeds words in multi-dimensional vectors. GloVe sought to compete with Word2Vec, and Socher and co-authors argued that for the same corpus, vocabulary, window size, and training time, GloVe consistently outperforms Word2Vec. The method used global word-word co-occurrence statistics to produce vector representations, which became widely adopted in [natural language processing](https://www.wikiprompt.org/wiki/natural-language-processing). Later, [Transformer](https://www.wikiprompt.org/wiki/transformer)-based models such as BERT improved upon GloVe, but the embeddings remained influential in the field.

## MetaMind and Salesforce

In 2014, Socher founded MetaMind, an AI and software company focused on [deep learning](https://www.wikiprompt.org/wiki/deep-learning) applications. The company developed systems for natural language understanding and image analysis, leveraging [neural networks](https://www.wikiprompt.org/wiki/neural-network) and [large language models](https://www.wikiprompt.org/wiki/large-language-model) in their early forms. Salesforce acquired MetaMind in 2016, and Socher served as Chief Scientist at Salesforce, where he led AI research and integration efforts. His work there contributed to Salesforce's Einstein platform, which applied [machine learning](https://www.wikiprompt.org/wiki/machine-learning) to customer relationship management.

## You.com and Recursive

In 2020, Socher left Salesforce to co-found You.com, an AI search infrastructure company. You.com aimed to provide a more personalized and privacy-focused search experience, integrating [generative AI](https://www.wikiprompt.org/wiki/generative-ai) and [large language models](https://www.wikiprompt.org/wiki/large-language-model) to answer queries. In September 2025, You.com raised $100 million in a Series C round led by Cox Enterprises, valuing the company at $1.5 billion. The platform gained attention for its use of [Transformer](https://www.wikiprompt.org/wiki/transformer)-based models and its focus on user control over search results.

In 2025, Socher co-founded Recursive, a company pursuing recursive self-improvement in AI, alongside seven other researchers. The company emerged from stealth in May 2026, having raised more than $650 million at a $4.65 billion valuation from investors including GV (formerly Google Ventures), Greycroft, Nvidia, and AMD. Recursive focuses on developing AI systems that can improve their own algorithms, a concept related to [AI](https://www.wikiprompt.org/wiki/artificial-intelligence) safety and [machine learning](https://www.wikiprompt.org/wiki/machine-learning) research.

## Publications and Later Work

In September 2026, PublicAffairs published Socher's first book, "The Eureka Machine: Why AI Is the Key to Unlocking a New Era of Scientific Discoveries", which argues that AI will accelerate the pace of science. The book draws on his experience in AI research and entrepreneurship, discussing how [deep learning](https://www.wikiprompt.org/wiki/deep-learning) and [generative AI](https://www.wikiprompt.org/wiki/generative-ai) can aid scientific discovery. Socher has also been an active investor through AIX Ventures, supporting early-stage AI startups. His work has been recognized with numerous awards, including being named to MIT Technology Review's 35 Innovators Under 35 in 2014.

Socher's contributions have shaped modern AI, from foundational datasets like ImageNet to practical applications in search and enterprise software. His current ventures reflect a continued focus on advancing AI capabilities and their societal impact.

---
Source: https://www.wikiprompt.org/wiki/richard-socher
License: CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/)
Last updated: 2026-09-09T01:58:45.80399+00:00