D

OpenAI's text-to-image generation model, integrated into ChatGPT and available via API. DALL-E 3 (2023) significantly improved prompt adherence and image quality. It is the most accessible AI image generator for non-technical users due to its integration directly into ChatGPT Plus.

A subset of machine learning that uses neural networks with many layers (hence 'deep') to learn complex patterns from large amounts of data. Deep learning is the technology behind modern AI breakthroughs in language understanding, image recognition, and video generation.

AI-generated synthetic media — typically video or audio — in which a person's likeness or voice is replaced or manipulated to appear as someone else. Deepfakes are created using generative AI and face-swapping algorithms. While they have legitimate uses in entertainment and education (such as AI avatars and dubbing), deepfakes pose significant risks for misinformation, fraud, and identity theft.

A Chinese AI company that released DeepSeek-V3 and DeepSeek R1 in late 2024/early 2025, causing significant disruption in the AI industry by demonstrating frontier-level performance at a fraction of the training cost of Western models. DeepSeek models are open-source and available via API, making them popular for cost-sensitive deployments.

An AI model where all parameters are used for every computation, as opposed to sparse models that activate only a subset. Dense models are simpler to train and reason about but are less efficient at scale. GPT-3 and early LLaMA models are examples of dense models.

An AI-powered audio and video editing tool that allows users to edit media by editing the transcript — deleting a word from the text removes it from the audio/video. Descript also features AI voice cloning, filler word removal, and screen recording, making it popular for podcasters and video creators.

A type of generative AI model that creates images, video, or audio by learning to reverse a noise-addition process. Diffusion models start with pure random noise and progressively denoise it into a coherent output guided by a text prompt. They power Midjourney, DALL-E, Stable Diffusion, Sora, and most modern AI image and video generators.

A virtual replica of a physical asset, process, or system that is continuously updated with real-world data and used for simulation, analysis, and optimization. In real estate and construction, digital twins of buildings allow owners to simulate energy usage, maintenance schedules, and renovation scenarios before making physical changes. AI enhances digital twins by enabling predictive analytics and autonomous optimization.