OpenAI AI: Timeline of Key Milestones
Founding and Early Years (Pre-2017)
Launch of OpenAI
Elon Musk and other prominent figures establish OpenAI, a non-profit research institute dedicated to ensuring artificial intelligence benefits humanity.
OpenAI Gym Released
This toolkit revolutionizes reinforcement learning research, providing a standardized environment for comparing and developing algorithms.
GPT Unveiled
The Generative Pre-trained Transformer model marks a significant leap in natural language processing, demonstrating remarkable text generation capabilities.
Growth and Advancements (2018-2020)
GPT-2 Sparks Debate
This powerful language model generates human-quality text, prompting concerns about potential misuse and ethical implications of such technology.
Microsoft Partners with OpenAI
This strategic partnership provides significant resources to accelerate AI research and development, aiming for beneficial outcomes.
GPT-3 Dominates the Scene
Unveiling the largest and most powerful language model to date, GPT-3 showcases exceptional capabilities in generating human-like text across various domains.
Expansion and Collaboration (2021-2022)
Expanding Research Focus
OpenAI diversifies its research areas, exploring reinforcement learning, robotics, and crucial topics like AI safety.
Global Collaboration
Partnerships with diverse organizations and researchers worldwide promote responsible AI development, fostering open discussion and collaboration.
Recent Developments (2022-Present)
ChatGPT Launches
OpenAI releases ChatGPT, a large language model chatbot, to the public.
ChatGPT Reaches 1 Million Users
Just four days after its launch, ChatGPT has already reached 1 million users. Two months later, the number of users exceeded 100 million.
Microsoft Invests $10 Billion in OpenAI
Microsoft invests $10 billion in OpenAI, the company behind ChatGPT.
OpenAI Launches ChatGPT Plus
OpenAI launches ChatGPT Plus, a subscription service that offers additional features and benefits for ChatGPT users.
OpenAI Releases GPT-4
OpenAI releases GPT-4, a new version of its large language model. ChatGPT Plus users are granted access to GPT-4.
ChatGPT Plus Users Gain Access to Third-Party Plugins and Browser Mode
ChatGPT Plus users gain access to third-party plugins and browser mode, which allows them to access the internet.
OpenAI Releases ChatGPT iOS App
OpenAI releases an iOS app for ChatGPT.
OpenAI Releases Code Interpreter Plugin for ChatGPT
OpenAI releases a code interpreter plugin for ChatGPT, which allows users to interpret code in natural language.
ChatGPT Gains Multimodal Capabilities
OpenAI announces that ChatGPT can now see, hear, and speak. Plus users can also upload images, and mobile app users can have conversations with the chatbot.
DALL-E 3 Integrated with ChatGPT
OpenAI’s latest image generation model, DALL-E 3, is integrated with ChatGPT Plus and ChatGPT Enterprise. The integration uses ChatGPT to help users write DALL-E prompts in a conversational way.
OpenAI Developer Conference
OpenAI holds its developer conference, where it announces GPTs, which allow users to customize their own GPTs. In the future, users will even be able to upload their own GPTs to the GPT store.
Sam Altman Removed as CEO
Sam Altman is removed as CEO of OpenAI by the board of directors.
Sam Altman Returns as CEO
Sam Altman returns to OpenAI as CEO.
OpenAI Releases Sora
OpenAI releases Sora, a text-to-video model. OpenAI also invites experts in biology, finance, healthcare, law, and sociology to join the Red Teaming Network. The network will work with OpenAI to rigorously evaluate and red team its AI models, in order to improve their safety.
OpenAI News
Google AI: Timeline of Key Milestones
Early Beginnings (Pre-2010)
Search Engine AI Foundations
Page & Brin lay groundwork for AI-powered search algorithms.
Machine Translation Takes Off
Google Translate launches, powered by machine learning.
Rise and Growth (2010-2016)
Deep Reinforcement Learning Joins the Team
Google acquires DeepMind.
AI Predicts Your Needs
Google Now launched, using AI for contextual information.
Deep Learning Brains
Google Brain project starts exploring its potential.
Democratizing AI
TensorFlow, an open-source machine learning framework, is released.
AI Organizes Your Photos
Google Photos uses AI for automatic image recognition and organization.
AI Conquers Go Champion
AlphaGo defeats Go world champion Lee Sedol, a major AI milestone.
Faster AI Training
Tensor Processing Units (TPUs) are introduced to accelerate AI deployments.
Formalization and Expansion (2017-2022)
Google AI Unites
Various AI efforts are brought under the Google AI umbrella.
Laying the Groundwork for Conversation
LaMDA, a language model for dialogue applications, is developed.
AI Makes Phone Calls
Google Duplex showcases the potential of AI for natural language interaction
Revolutionizing NLP
BERT, a Transformer-based language model, improves natural language processing.
Expanding Scope and Resources
Google AI becomes part of Google Research.
Understanding Beyond Text
MUM, a multimodal understanding model, is introduced.
Conversational AI Enters Beta
Bard, powered by LaMDA, begins beta testing.
Recent Developments (2023-Present)
Bard Unveiled
Feeling the heat from ChatGPT and Microsoft, Google officially launches its chatbot, Bard.
Bard Demo Mishap
Google held a news conference for Bard, but the promotional video aired that day contained an error, providing an incorrect answer to an astronomical question.
The market perceived this as Google rushing Bard to market in panic, leading to a sudden drop in Google’s market value by over a trillion dollars overnight. Subsequently, Google resolved to make a strong comeback.
PaLM-E Revealed
Google unveils PaLM-E, the largest visual language model ever created, boasting a whopping 562 billion parameters. This impressive model integrates visual and language capabilities, allowing it to control robots with remarkable precision.
PaLM 2 Enters the Arena
Google enters the ring against GPT-4 with the release of PaLM 2, a 340 billion-parameter model. Despite its smaller size, Google claims PaLM 2 can tackle complex tasks with greater efficiency, marking another step forward in the AI race.
Duet AI Preview Launch
Google Cloud introduces Duet AI, an AI-powered collaborator, in Preview. This marks the initial rollout of Duet AI, offering users a glimpse into its capabilities.
Gemini Trio Unleashed
The highly anticipated multimodal AI model, Gemini, finally arrives in three versions: Nano, Pro, and Ultra.
Pro targets GPT-3.5, while Ultra takes aim at OpenAI’s GPT-4. Proclaimed by CEO Sundar Pichai as Google’s most powerful AI to date, Gemini can process information beyond text, incorporating images, videos, and audio.
Demo videos showcase its potential, combining vision and hearing in impressive ways.
Imagen 2 Enhances Image Generation
Google DeepMind releases Imagen 2, an enhanced text-to-image model. This model boasts significant improvements in image quality and new features like text rendering, making it suitable for commercial applications and logo generation.
MedLM Raises Medical AI Standards
Google continues its push in medical AI with MedLM, a model based on Med-PaLM 2. This model achieves an impressive 86.5% accuracy on the USMLE medical licensing exam, equivalent to an expert doctor according to Google.
Duet AI Generally Available
Duet AI officially becomes Generally Available, offering developers an AI-powered collaborator with improved model accuracy and comprehensive responses. Key features include refined code generation in various IDEs, enhanced filtering and sorting, and improved logic for code completion.
VideoPoet Revolutionizes Video Generation
Google brings video generation to the table with VideoPoet. This model sets itself apart by adding audio effects, allowing interactive editing, and most importantly, extending the generated video length beyond the usual short clips.
By cleverly using AI to predict the next second based on the previous one, VideoPoet opens new possibilities for video creation.
Duet AI Admin Controls
Admin controls for Duet AI subscriptions become available in the Google Cloud console, providing users with more flexibility to manage their subscriptions and access Duet AI features.
Gemini News
Duet AI News
Microsoft AI: Timeline of Key Milestones
Early Research (1990s to 2010s)
Founding AI Research Groups
Microsoft Research begins investing in AI with dedicated groups focused on language understanding, machine learning, and computer vision.
Speech API for Developers
Microsoft releases the Speech API, allowing developers to incorporate speech recognition capabilities into their applications.
Kinect Launch (AI-powered Motion Capture)
Microsoft launches the Kinect for Xbox 360, featuring innovative AI-powered motion capture technology.
Bing Search with AI Capabilities
Microsoft introduces AI features to Bing, enhancing its search engine capabilities with improved natural language processing and understanding.
Minecraft Acquisition (AI In-Game Experiences)
Microsoft acquires Minecraft, opening doors for using AI to create unique and engaging in-game experiences.
Acceleration and Partnership (2010s to 2020s)
AI for Good Initiative (Responsible Development)
Microsoft establishes the AI for Good initiative, demonstrating their commitment to ethically responsible development and application of AI technology.
OpenAI Partnership (Cutting-edge AI Research)
Microsoft forms a key partnership with OpenAI, gaining access to their advanced AI research and contributing to further innovation.
Cortana Virtual Assistant Launch
Microsoft introduces Cortana, an AI-powered virtual assistant designed to help users with tasks and information access.
Azure Cognitive Services for Developers
Microsoft unveils Azure Cognitive Services, a suite of AI tools providing developers with functionalities like computer vision, speech recognition, and language translation for their applications.
Azure Machine Learning Service
Microsoft releases the Azure Machine Learning service, empowering users to build, deploy, and manage AI models for various purposes.
Project Turing Announcement (Large Language Models)
Microsoft announces Project Turing, showcasing their ambitious commitment to developing and exploring large language models similar to GPT-3.
Recent Developments (2023-Present)
Powering Up with OpenAI
Microsoft invests $100 billion in OpenAI, acquiring a 49% stake. This investment gives Microsoft access to OpenAI’s GPT-4 technology, which is a powerful language model that can generate text, translate languages, write different kinds of creative content, and answer your questions in an informative way.
Unveiling New Bing (Bing Chat)
Microsoft releases New Bing (Bing Chat), which integrates GPT-4. New Bing is a search engine that uses GPT-4 to provide more natural and informative search results.
Introducing Microsoft 365 Copilot
Microsoft releases Microsoft 365 Copilot, which integrates GPT-4 into Microsoft 365 applications. Copilot can help users with a variety of tasks, such as writing emails, creating documents, and managing their calendars.
Copilot Takes Flight
Microsoft announces that Bing Chat has been renamed Copilot and is now available for free with GPT-4 and DALL-E 3 capabilities. Copilot will also be embedded in all Microsoft applications. Microsoft also releases Copilot Studio, which allows users to customize GPT, create plugins, and customize workflows.
Expanding Copilot’s Reach
Microsoft Introduces Copilot Pro, a new subscription tier for individuals with advanced features. Copilot is now available for businesses of all sizes, including small and medium businesses.
Meta AI: Timeline of Key Milestones
Foundational Years (2013-2017)
Founding FAIR
Facebook announces the creation of the Facebook AI Research Lab (FAIR), led by Yann LeCun, a pioneer in deep learning.
Vladimir Vapnik Joins FAIR
Renowned computer scientist Vladimir Vapnik, a leading figure in statistical learning theory, joins FAIR.
Paris Lab & Company Rebranding
FAIR opens a research center in Paris, and Facebook rebrands itself as Meta Platforms.
Partnership on AI
FAIR collaborates with other companies to establish the Partnership on Artificial Intelligence to Benefit People and Society.
Meta AI & PyTorch Release
FAIR renames itself Meta AI and releases PyTorch, an open-source machine learning framework.
Growth and Innovation (2018-2022)
Staff Growth & Language Translation
Meta AI doubles its staff size and unveils a system capable of translating between written languages.
Top Ranking in AI Research
Meta AI is ranked 8th in the AI Research Rankings.
Protein Structure Prediction
Meta AI achieves a breakthrough by predicting the 3D structure of 600 million proteins in just two weeks.
Recent Developments (2023-Present)
Meta Releases Open-Source Large Model LLaMA
Meta releases LLaMA, an open-source large language model with 65 billion parameters.
Meta Releases Open-Source Large Model LLaMA 2
Meta releases LLaMA 2, an open-source large language model with 700 billion parameters. This move garners positive attention, with many AI researchers commenting that “Meta became cool again.” This shift has a positive impact on the company’s image and its ability to attract talent.
Meta’s commitment to open-source AI was foreshadowed in the book “Genius Makers,” which mentions that when Mark Zuckerberg invited Yann LeCun, one of the “three giants of deep learning,” to join Facebook and lead AI development, LeCun’s condition was that Meta’s AI development must be open-source. Zuckerberg agreed, leading to the current LLaMA open-source roadmap.
