Releases

Google researchers introduce the Transformer architecture, replacing recurrence with attention and reshaping modern sequence modeling.

OpenAI released Baselines, open-source high-quality implementations of reinforcement-learning algorithms.

At Google I/O, the company frames its strategic transition from mobile-first to AI-first products and computing.

Google announces its second-generation TPU and plans to offer Cloud TPUs to Google Cloud customers.

NVIDIA introduced its GPU Cloud catalog, later branded NGC, for optimized deep-learning containers and software.

NVIDIA introduced the Volta architecture and V100 GPU, including Tensor Cores built for deep learning.

Baidu announced its open autonomous-driving platform.

Facebook released the Caffe2 deep-learning framework for scalable cross-platform training and deployment.

Facebook AI Research released FAISS for efficient similarity search and clustering of dense vectors.

Hugging Face raised approximately $1.2 million in early angel funding to develop its conversational AI startup.

Baidu introduced its voice-assistant experience powered by DuerOS.

Baidu introduced its conversational AI operating system for devices and services.

IBM commercialized enterprise conversational-AI tooling that evolved into watsonx Assistant.

IBM launched an AI-powered enterprise search and document-understanding service.

Scale began building distributed contributor platforms and training operations to support large-scale annotation, expert-data, and evaluation programs.

Scale launched Remotasks to train and coordinate distributed contributors performing data annotation and review tasks.

Databricks raised a $60 million Series C led by New Enterprise Associates.

Amazon unveiled a checkout-free store using computer vision, sensor fusion and deep learning.

OpenAI released Universe, a software platform for training intelligent agents across games, websites, and applications.

AWS introduced a managed text-to-speech service with natural-sounding voices.

AWS introduced the conversational AI service built from the speech recognition and language understanding technology behind Alexa.

AWS introduced managed image analysis for object, scene and face detection.

AWS selected MXNet as its preferred deep-learning framework and committed engineering investment to the open-source project.

Microsoft creates the AI and Research Group, combining research, information-platform and AI product teams under Harry Shum.

Oracle introduced the infrastructure platform that became OCI, later forming the base for its managed AI services and GPU cloud.

DeepMind introduces WaveNet, a neural generative model for raw audio waveforms that substantially improves speech synthesis.

Baidu launched its core AI technology engine and open-capability platform.

Baidu opened PaddlePaddle to the global developer community for deep learning, vision, speech and NLP.

Facebook AI Research released PyTorch, a Python-first tensor and dynamic neural-network framework.

Facebook AI Research released fastText, a library for efficient text classification and word representations.

Scale launched the data infrastructure that became its core Data Engine, combining APIs, machine assistance, and human review to create high-quality training data.

Scale launched an API that let developers submit human-powered data tasks with one line of code, establishing the foundation of its AI data business.

Alexandr Wang and Lucy Guo founded Scale in San Francisco and entered Y Combinator to build an API for routing human work, which evolved into AI data infrastructure.

Google publicly reveals the TPU, a custom accelerator designed to speed machine-learning workloads in its data centers.

Amazon open-sourced its Deep Scalable Sparse Tensor Network Engine for recommendation workloads.

Facebook introduced FBLearner Flow, its internal platform for managing the full lifecycle of machine-learning models.

OpenAI released the public beta of Gym, a toolkit for developing and comparing reinforcement-learning algorithms.

The Pascal-based Tesla P100 accelerator debuted alongside DGX-1 for deep learning and HPC.

NVIDIA unveiled DGX-1, its first integrated deep-learning supercomputer.

Microsoft launches Cognitive Services APIs for adding vision, speech, language, knowledge and search capabilities to applications.

Microsoft takes the Tay social chatbot offline after coordinated abuse caused it to produce offensive content, prompting a public review of safeguards and social-AI deployment.

AlphaGo wins its landmark five-game match against 18-time world champion Lee Sedol by four games to one.

Microsoft releases its Computational Network Toolkit under an open-source license for deep neural-network training.

Replit launched its browser-based development workspace.

Replit was founded to make programming accessible through an online, collaborative computing environment.

Shield AI introduced Nova for autonomous reconnaissance in buildings and GPS-denied environments.

Palantir established the Ontology as Foundry's semantic and operational layer.

Data Connection became Foundry's connector and ingestion layer.

Palantir introduced Foundry as its enterprise data operations platform.

Co-founder Ali Ghodsi became chief executive officer of Databricks.