Releases

Diffusers added first-class support for Stable Diffusion, enabling local and hosted generation workflows around the newly released open model.

Hugging Face launched its enterprise platform around Private Hub, secure collaboration, governance, deployment, and organizational support.

Hugging Face launched Private Hub as a secure and compliant platform for organizations to collaborate on models, datasets, and applications.

Amazon researchers published a 20-billion-parameter multilingual sequence-to-sequence teacher model.

Baidu launched an ERNIE-powered AI art and creative-generation service.

Hugging Face released Transformer Reinforcement Learning tooling for post-training language models with supervised, preference, and reinforcement-learning methods.

Aravind Srinivas, Denis Yarats, Johnny Ho and Andy Konwinski founded Perplexity in San Francisco.

The AlphaFold database expands to predicted structures for nearly every catalogued protein known to science.

Baidu unveiled its sixth-generation purpose-built autonomous vehicle.

Hugging Face launched Diffusers, an open-source library providing modular pipelines for diffusion-based generative models.

Meta introduced Make-A-Scene, a multimodal generative method that gives creators scene-level control over images.

The BigScience workshop released BLOOM, a 176-billion-parameter multilingual language model trained on the Jean Zay supercomputer.

Meta released NLLB-200, a machine-translation model covering 200 languages.

AWS introduced its generative-AI coding companion in preview.

Amazon unveiled its first fully autonomous mobile robot for safely operating around employees.

Microsoft publishes the second version of its Responsible AI Standard and supporting transparency materials.

GitHub Copilot becomes generally available to individual developers as a subscription product.

The Cerner acquisition became the foundation for Oracle Health's clinical and data-intelligence AI portfolio.

Oracle completed its acquisition of Cerner to create a major health technology and clinical-data business.

Hugging Face launched Evaluate as a unified framework for model and dataset evaluation.

Microsoft releases the Responsible AI Dashboard, combining fairness, interpretability, error analysis and causal tools in one interface.

Google Research introduces Imagen, a text-to-image diffusion model emphasizing photorealism and language understanding.

Hugging Face raised $100 million in Series C funding led by Lux Capital at a reported $2 billion post-money valuation.

Meta released the Open Pretrained Transformer family and provided OPT-175B access to researchers.

Baidu released an inference and deployment toolkit for AI models.

Anthropic raised $580 million to build steerable, interpretable, and robust AI systems and expand its research infrastructure.

Modal raised $7 million in seed financing to build its serverless cloud platform.

OpenAI introduced DALL-E 2 with higher-resolution image generation and editing capabilities.

IBM introduced z16 with Telum-powered real-time AI inference for mission-critical transactions.

Google Research introduces the 540-billion-parameter Pathways Language Model and reports strong few-shot performance across hundreds of tasks.

NVIDIA expanded TAO Toolkit integration for fine-tuning and optimizing pretrained production AI models.

NVIDIA made its end-to-end GPU-accelerated recommender AI framework generally available.

NVIDIA made its multilingual speech AI SDK generally available.

NVIDIA announced its fourth-generation DGX system powered by eight H100 GPUs.

NVIDIA launched Jetson AGX Orin for next-generation robotics and autonomous machines.

NVIDIA announced the Grace Hopper Superchip, pairing a Grace CPU with a Hopper GPU using coherent memory.

NVIDIA announced the Hopper architecture and H100 GPU with a Transformer Engine for large AI models.

Microsoft completes its acquisition of Nuance, adding clinical speech recognition, ambient intelligence and conversational AI to its portfolio.

Hugging Face introduced browser previews and automatic processing for datasets hosted on the Hub.

Hugging Face introduced SafeTensors as a fast tensor serialization format designed to avoid arbitrary-code execution.

Cohere raised $125 million in a Series B led by Tiger Global to expand its enterprise language-model platform.

Scale expanded its data engine beyond perception data into expert demonstrations, preference feedback, red teaming, and reinforcement learning from human feedback for language models.

OpenAI introduced instruction-following models trained with human feedback.

Meta open-sourced TorchRec, a PyTorch library for large-scale recommendation systems.

OpenAI released a new endpoint and model families for text and code embeddings.

The U.S. government awarded Scale a contract vehicle worth up to $249 million to make its AI data and development capabilities available across federal agencies.

Meta announced the AI Research SuperCluster, a large AI supercomputer designed to train advanced models.

IBM agreed to sell healthcare data and analytics assets from Watson Health to Francisco Partners.

Michael Truell, Sualeh Asif, Arvid Lunnemark, and Aman Sanger founded Anysphere to build an AI-native programming environment.

Palantir introduced FedStart to help software vendors reach government customers through accredited infrastructure and deployment support.

Cohere began offering enterprise model deployment within customer-controlled cloud and on-premises environments.

Cohere introduced its Embed model family for semantic representation and retrieval.

Cohere launched a free educational curriculum covering large language models, embeddings and semantic search.

Cohere established its open-science research initiative, now known as Cohere Labs, to support collaborative machine-learning research.

Hugging Face acquired Gradio, bringing its open-source interface framework and team into the company while keeping the project open source.

Palantir unveiled MetaConstellation, a satellite-tasking and geospatial intelligence capability that coordinates space-based data with AI.

AWS introduced a no-code visual interface for generating machine-learning predictions.

Microsoft releases DeBERTa V3 with replaced-token detection and gradient-disentangled embedding sharing.

Oracle released managed time-series forecasting for business and operational signals.

Oracle released managed multivariate time-series anomaly detection.

Oracle released managed image analysis, object detection and custom vision models.

Oracle released a managed natural-language processing API for classification, entity extraction and sentiment.

Oracle released managed automatic speech recognition and transcription.

Oracle introduced a portfolio of managed AI APIs spanning language, speech, vision, anomaly detection and forecasting.

CoreWeave raised $50 million from Magnetar Capital to scale its specialized NVIDIA GPU cloud infrastructure.

Scale acquired Berlin-based SiaSearch and folded its petabyte-scale sensor-data search technology and team into Scale Nucleus.

Microsoft introduces Azure OpenAI Service, providing managed access to OpenAI models through Azure's enterprise cloud controls.

Facebook renamed the parent company Meta Platforms to reflect its long-term focus on building the metaverse.

Facebook AI and university partners introduced Ego4D, a large egocentric video dataset and benchmark suite.

Scale moved Rapid into general availability for fast labeling experiments without data minimums.

Hugging Face launched Spaces with native Gradio support, allowing community members to host interactive machine-learning demos on the Hub.

Databricks closed a $1.6 billion Series H with participation from AWS, CapitalG and Microsoft at a $38 billion post-money valuation.

Cohere's hosted API made semantic text representations available to enterprise developers.

Cohere provided a browser playground for testing its hosted language models and generation settings.

Cohere's initial commercial platform included text classification for enterprise NLP workflows.

Index Ventures led Cohere's $40 million Series A, with Section 32, Radical Ventures and AI researchers participating.

Cohere made its hosted language-model platform available to businesses alongside its Series A financing.

Hugging Face launched Optimum to optimize Transformers models for specialized hardware and inference runtimes.

NVIDIA made its AI Enterprise software suite generally available.

The Telum announcement established a new generation of IBM infrastructure with integrated AI acceleration.

IBM unveiled Telum with on-chip AI acceleration for high-volume transactional inference.

Baidu said its second-generation Kunlun AI chip had entered mass production.

OpenAI introduced Codex, a descendant of GPT-3 that translates natural language into code.

Scale established a dedicated St. Louis operations center to develop skilled U.S. data and evaluation workforces for government programs.

DeepMind and EMBL-EBI launch a freely accessible database of AlphaFold protein structure predictions.

DeepMind publishes AlphaFold 2's methodology and releases source code, expanding access to high-accuracy protein prediction.

Baidu published its unified framework for knowledge-enhanced language understanding and generation.

AWS founder Andy Jassy succeeded Jeff Bezos as Amazon's chief executive officer.

GitHub launches Copilot in technical preview as an AI pair programmer that suggests code inside the editor.

NVIDIA made Fleet Command generally available for centrally deploying and managing edge AI applications.

NVIDIA launched a hosted enterprise program offering immediate access to its end-to-end AI infrastructure and software stack.

Hugging Face launched AutoTrain to automate model training and fine-tuning from user-provided datasets.

Anthropic raised $124 million to pursue computationally intensive research into reliable, interpretable, and steerable AI systems.

Databricks introduced a managed Feature Store integrated with Delta Lake, MLflow and model serving.

Databricks introduced Unity Catalog as a unified governance layer for data and AI assets across clouds.

Databricks introduced AutoML to automatically train models while generating editable notebooks.

Facebook AI introduced unsupervised speech recognition using unpaired speech audio and text.

Google announces fourth-generation TPU pods for large-scale machine-learning training.

Google Cloud launches Vertex AI as a unified platform for building, deploying and operating ML models.

Google introduces LaMDA as a conversational language model designed for open-ended dialogue.