Generative modeling
Generative modeling is a machine-learning approach that learns patterns or a data distribution to generate new samples resembling the data.
Starting Data Analytics with Generative AI and Python
A machine-learning model that generates new content, such as text, code, or explanations, from learned patterns and user instructions.
- How it is used
- Supports analytical planning, coding, text analysis, and interpretation while leaving data understanding and decisions with the analyst.
- How deep it goes
- Develops a prompt–run–inspect–revise workflow and practical safeguards, while deliberately deferring model architecture and advanced API integration.
- Examples
- Uses product-category sales code, regional delivery-time charts, and customer-review analysis to show both useful suggestions and failures.
- What you can do after
- Apply it via ChatGPT or Gemini to draft and refine analyses, then execute code and verify claims against data.
Read how Starting Data Analytics with Generative AI and Python explains generative modeling →
Introduction to Generative AI, Second Edition
Generative AI is the area of artificial intelligence focused on models that produce new content such as text, images, audio, or code.
- How it is used
- Frames content-generating AI as a tool for creative, educational, and professional work whose outputs require human judgment.
- How deep it goes
- Surveys applications and risks at a high level, without giving a procedure for building generative models.
- Examples
- Uses ChatGPT conversations, AI-generated ketchup images, and a legal brief containing invented cases.
- What you can do after
- Evaluate/understand when generated content can assist a task and when claims, privacy, or professional consequences demand extra scrutiny.
Read how Introduction to Generative AI, Second Edition explains generative modeling →
Generative AI in Action
Generative AI is an artificial-intelligence approach that produces new content or processes rather than merely classifying or predicting properties of existing inputs.
- How it is used
- Positions generative modeling as a way to create content and support enterprise workflows beyond predefined classification or prediction tasks.
- How deep it goes
- Explains learned patterns, prompts, applications, and adoption considerations, while deferring detailed machine-learning and deep-learning mechanisms.
- Examples
- Uses GPT-4 email drafting and entity extraction, alongside enterprise chatbot and industry use cases.
- What you can do after
- Evaluate/understand it for an enterprise use case, then plan a small pilot with data, integration, governance, and success criteria.
Read how Generative AI in Action explains generative modeling →
Coding with AI
Artificial intelligence that produces new content, such as text, source code, documentation, or tests, from learned patterns and input instructions.
- How it is used
- Supports coding, bug detection, documentation, testing, planning, and design throughout an AI-assisted development workflow.
- How deep it goes
- Explains learned-pattern prediction and introduces LLM Transformer architecture, while deferring advanced prompting, context management, and testing practice.
- Examples
- Uses a HAM radio practice-exam Flask application, generated stubs and classes, and AI-created test suites as practical artifacts.
- What you can do after
- Apply it via ChatGPT, Gemini, GitHub Copilot, Tabnine, or Blackbox AI while validating and revising generated work.
Build Financial Software with Generative AI (From Scratch)
Generative artificial intelligence, or generative AI, produces new content such as text, code, images, or other data from learned patterns and input.
- How it is used
- Supports full-stack ACH dashboard work by generating code, troubleshooting failures, recalling syntax, and creating test data.
- How deep it goes
- Defines generative AI and LLMs, then demonstrates prompting, code refinement, debugging, syntax assistance, privacy, and verification without comprehensive AI coverage.
- Examples
- Uses balance-formatting Python code, a transaction-list defect with unit tests, an HTML5 D3.js bar chart, ACH-file parsing, and company test data.
- What you can do after
- Apply it via ChatGPT or GitHub Copilot to draft, refine, and investigate code while verifying results and protecting sensitive inputs.
Read how Build Financial Software with Generative AI (From Scratch) explains generative modeling →
Effective Conversational AI
An AI model that creates new content, such as text, images, audio, or code, from learned patterns or conditioning information.
- How it is used
- Supports customer-facing answers and build-time improvements alongside controlled conversational AI.
- How deep it goes
- Explains text generation through next-word prediction and develops practical selection, guardrail, and assessment criteria without deep implementation code.
- Examples
- Uses varied bakery recipe requests, password-reset training utterances, process flows, and conversation transcripts as practical cases.
- What you can do after
- Evaluate/understand it by choosing controlled or generated responses, testing correctness, and setting guardrails proportionate to the risk of errors.
Read how Effective Conversational AI explains generative modeling →
First-Party Data Activation
Artificial intelligence that generates new content, such as text, images, audio, video, or code, from learned patterns.
- How it is used
- Supports marketing ideation, targeting, personalization, creative production, and interpretation of results using first-party data.
- How deep it goes
- Defines generated outputs against predictive scores and develops practical applications and adoption considerations, without teaching model implementation.
- Examples
- Uses tailored retailer emails, conversational audience suggestions, and natural-language business intelligence as concrete applications.
- What you can do after
- Evaluate/understand where generative marketing can assist campaign work and what data, measurement, and governance conditions affect adoption.
Read how First-Party Data Activation explains generative modeling →
Data Without Labels
Generative AI is artificial intelligence that generates new content, such as text, images, or video, from learned patterns and an input prompt or other condition.
- How it is used
- Positions generative modeling as a way to create new content and synthetic data rather than only predict or classify.
- How deep it goes
- Defines the data-distribution objective and gives an integration checklist, but does not work through a generative-model implementation.
- Examples
- Uses text drafting, translation, chatbots, image generation, and synthetic data to illustrate possible outputs.
- What you can do after
- Evaluate/understand it by distinguishing generation from classification and planning data preparation, model selection, and validation for a use case.
Build AI-Enhanced Web Apps
Generative AI is an artificial-intelligence paradigm that generates new content from learned patterns rather than only classifying inputs or following predefined rules.
- How it is used
- Enables web applications to generate personalized content and conversational responses from user input instead of relying solely on predefined content.
- How deep it goes
- Explains pattern-based generation, LLM prediction, application components, and validation concerns without working through model-training internals.
- Examples
- Uses marketing image and copy generation, customer-support responses, and a mock interview application as product examples.
- What you can do after
- Apply it via React, Next.js, and the Vercel AI SDK to add model-backed features while checking generated outputs.
Read how Build AI-Enhanced Web Apps explains generative modeling →
AI for Everyday IT
Generative AI is a class of artificial intelligence systems that produces new content, such as text, images, code, or structured data, from learned patterns and inputs.
- How it is used
- Improves IT workflows by generating content, analyzing information, supporting communication, and assisting documentation.
- How deep it goes
- Defines learned-pattern generation and practical uses, but stops short of deep model theory, mathematics, and internal mechanics.
- Examples
- Uses PDF extraction, log analysis, chatbot comparisons, workplace messaging, application integration, and MySQL disaster-recovery documentation.
- What you can do after
- Apply generative AI to routine IT tasks while validating outputs and protecting organizational data.
Inside AI
Generative artificial intelligence is an approach in which models produce new content such as text, images, audio, or code from learned patterns.
- How it is used
- Uses generative modeling to explain content creation and accessible AI tools such as ChatGPT.
- How deep it goes
- Defines generation through learned data patterns and notes its statistical roots, but does not explain model internals.
- Examples
- Illustrates uses with customer-support responses and code generation, and risk with fabricated legal citations.
- What you can do after
- Evaluate/understand when generated content may help and why factual, safety, and ownership checks remain necessary.
Effective Data Analysis
A class of artificial-intelligence systems that generates new content such as text, code, images, audio, or video from learned patterns and inputs.
- How it is used
- Positions generated content as an aid to analytics workflows and organizational decisions, not a replacement for analysts.
- How deep it goes
- Defines generative AI and foundation models, then develops prompt refinement and evaluation rather than model-training internals.
- Examples
- Uses customer-login prompts and a rat-sightings prediction exercise to test responses, code, metrics, and recommendations.
- What you can do after
- Practice/perform iterative chatbot prompting and evaluate its analysis against an existing model and stakeholders’ needs.
Read how Effective Data Analysis explains generative modeling →
Writing for Developers: Blogs that get read
Artificial intelligence that generates new content such as text, images, audio, or code from learned patterns.
- How it is used
- Supports rapid gap-finding and grammar checking during technical writing, while preserving human control over content and revisions.
- How deep it goes
- Explains practical prompts and review safeguards, but stops short of describing model internals or reliable automated judgment.
- Examples
- Uses engineering-blog gap prompts, grammar-rule questions, full-article checks, and automated detection of article and its/it’s errors.
- What you can do after
- Apply it via an AI chatbot to surface possible omissions or language errors, then diagnose each suggestion before changing the draft.
Read how Writing for Developers: Blogs that get read explains generative modeling →
The Art of AI Product Development
Artificial intelligence that generates new content such as text, images, audio, or code from learned patterns.
- How it is used
- Positions generative modeling as a solution choice when the task calls for new content rather than prediction or workflow execution.
- How deep it goes
- Defines generation through learned patterns and contrasts AI types, without detailing training or generation mechanics.
- Examples
- Uses GPT-4o sentiment classification for movie reviews as a case where logistic regression may be cheaper and easier to host privately.
- What you can do after
- Evaluate/understand it as one solution option, checking whether a simpler predictive model better fits an analytical task.
Read how The Art of AI Product Development explains generative modeling →
Deep Learning with R, Third Edition
Generative AI comprises computational models that produce new content such as text, images, audio, or code from learned patterns.
- How it is used
- Uses generation to connect learned patterns with prompt-driven chatbots, image tools, and creative assistance.
- How deep it goes
- Explains reconstruction, sampling, and prompting conceptually; sketches miniature GPT and Transformer chatbot projects without detailing their implementations.
- Examples
- Uses next-word prediction, image denoising, ChatGPT, and a Midjourney image generated from a descriptive prompt.
- What you can do after
- Evaluate/understand generated content as sampled from learned structure and steerable by people, rather than as independent artistic intent.
Read how Deep Learning with R, Third Edition explains generative modeling →
Data-Driven Cybersecurity
Generative AI is artificial intelligence that creates new content such as text, images, audio, or code from learned patterns.
- How it is used
- Uses generated data and reports to support cybersecurity metrics analysis, scenario testing, and decision-making.
- How deep it goes
- Works through a prompt-to-chart data pipeline and incident examples, but does not develop model-training internals.
- Examples
- Uses fictitious failed-login records, a failure-reason bar chart, and an incident-response table.
- What you can do after
- Apply it via a local language model to generate test data and draft analyses, then validate findings before acting.
Read how Data-Driven Cybersecurity explains generative modeling →
Graph Neural Networks in Action
A deep generative model is a generative model that uses neural networks to learn and generate data distributions.
- How it is used
- Uses learned data distributions to sample synthetic examples and, in graph settings, support reconstruction and edge prediction.
- How deep it goes
- Explains the probability distinction and graph autoencoder mechanisms, then compares architecture trade-offs without a rigorous derivation.
- Examples
- Uses boat and airplane images, synthetic faces, candidate drugs, and graph edge prediction as examples.
- What you can do after
- Apply it via PyTorch Geometric to build graph autoencoders for edge prediction and choose between GAE and VGAE for graph tasks.
Read how Graph Neural Networks in Action explains generative modeling →
How Large Language Models Work
Generative AI/ML is the area of artificial intelligence and machine learning concerned with producing new content or outputs.
- How it is used
- Frames content generation as a capability for evaluating applications, limitations, and deployment decisions across AI systems.
- How deep it goes
- Defines generated outputs and sketches input-output behavior, but stops short of implementation details or building models from scratch.
- Examples
- Uses haiku generation, text summarization, rewriting, detail addition, completion, and multimodal input-output transformations.
- What you can do after
- Evaluate/understand what generative systems can produce, where they may fit, and when reliability or cost makes them unsuitable.
Read how How Large Language Models Work explains generative modeling →
Outlier Detection in Python
A generative model represents a process or probability distribution capable of assigning likelihoods to observations or generating new observations.
- How it is used
- Uses probability models to assess typicality, generate synthetic observations, and support outlier detection.
- How deep it goes
- Explains likelihood fitting, sampling, covariance assumptions, and reconstruction-based variants, but does not derive their mathematics.
- Examples
- Uses baseball records, Gaussian clusters, synthetic samples, covariance expansion, and image outlier-detection methods.
- What you can do after
- Evaluate typicality, generate synthetic tabular data with GMMs, and compare generative detectors using synthetic data or pseudo-ground truth.
Read how Outlier Detection in Python explains generative modeling →
Learn Generative AI with PyTorch
A class of AI systems that create new content such as text, images, code, audio, music, or video rather than only classifying existing inputs.
- How it is used
- Explains how AI systems produce novel content instead of only recognizing or labeling existing inputs.
- How deep it goes
- Defines outputs, contrasts probability objectives, and sketches GANs and Transformers without developing training procedures or derivations.
- Examples
- Uses dog-and-cat image synthesis, text-to-image systems, ChatGPT, generated music, legal documents, code, audio, and video.
- What you can do after
- Evaluate/understand how generative and discriminative models differ and identify major generative model families and applications.
Read how Learn Generative AI with PyTorch explains generative modeling →
Acing the CCNA Exam: Volume 2 Advanced Networking and Security
Generative AI is artificial intelligence that produces new content such as text, code, configurations, or designs from learned patterns and supplied requirements.
- How it is used
- Supports network automation by generating operational artifacts and assisting intent-based networking.
- How deep it goes
- Defines learned-pattern generation and contrasts prediction, but stops short of explaining model internals or implementation procedures.
- Examples
- Uses ChatGPT, Gemini, Midjourney, DALL-E, scripts, diagrams, documentation, and plug-and-play deployments as examples.
- What you can do after
- Evaluate generative AI as an aid for producing network scripts, diagrams, documentation, deployments, and intent-based configurations.
Investing for Programmers
Artificial intelligence that generates new content or information, including text, images, code, or synthesized insights.
- How it is used
- Supports financial research by generating summaries, company insights, sentiment analysis, and scenario-oriented analysis.
- How deep it goes
- Defines GenAI, contrasts it with discriminative ML, and demonstrates applications while deferring transformer and attention internals.
- Examples
- Uses earnings reports, news articles, social media sentiment, management questions, investment scorecards, and Python LLM integrations.
- What you can do after
- Apply it via Python and LLM platforms to research companies, generate insights, and identify features for focused ML models.
Read how Investing for Programmers explains generative modeling →
Deep Learning with Python, Third Edition
AI systems that generate new text, images, or other content in response to prompts by modeling the distribution of data.
- How it is used
- Uses learned data distributions to generate content from prompts across text, images, video, and sound applications.
- How deep it goes
- Explains self-supervised objectives, parameters, representations, and sampling, but stops short of specific model architectures or implementation code.
- Examples
- Uses ChatGPT, Gemini, Claude, and Midjourney alongside next-word prediction and image denoising as concrete examples.
- What you can do after
- Evaluate generative AI applications by relating their prompted outputs to self-supervised training and learned model representations.
Read how Deep Learning with Python, Third Edition explains generative modeling →
The Complete Obsolete Guide to Generative AI
A machine-learning model that produces new content resembling patterns learned from its training data.
- How it is used
- Applies generative AI tools to practical work such as analysis, summarization, fine-tuning, and content generation.
- How deep it goes
- Introduces model classes and prompt controls while minimizing under-the-hood theory and technical background.
- Examples
- Illustrates iterative code generation and model-to-model feedback, alongside analysis, summarization, audio, and video use cases.
- What you can do after
- Apply it via interactive tools and APIs to configure prompts and automate model interactions.
Read how The Complete Obsolete Guide to Generative AI explains generative modeling →
Machine Learning for Tabular Data: XGBoost, Deep Learning, and AI
Artificial intelligence that generates new content, such as text, images, code, or structured outputs, from learned patterns.
- How it is used
- Supports text processing, tabular-data analysis, feature engineering, coding, visualization, reporting, and machine-learning pipeline work.
- How deep it goes
- Explains practical capabilities and workflow applications, but stops short of detailing model architecture, training, or generation algorithms.
- Examples
- Uses textual table fields, CSV analysis, generated Python functions, visualizations, reports, pipeline code, documentation, and summarized logs.
- What you can do after
- Apply it via ChatGPT’s Advanced Data Analytics API or Gemini for Google Cloud to analyze data and support pipeline development.
Evolutionary Deep Learning: Genetic algorithms and neural networks
Generative modeling is the machine-learning task of modeling a data-generating process so that its outputs can be replicated or newly generated.
- How it is used
- Frames image replication as a generative-modeling problem and connects it to later generative deep-learning applications.
- How deep it goes
- Implements encoding, rendering, fitness evaluation, and genetic search, but stops short of developing GAN internals.
- Examples
- Uses EvoLisa, polygon brushes, target and random renderings, pixel-error fitness, and alternative rectangle or polyfill formats.
- What you can do after
- Build it from scratch as a DEAP genetic-algorithm pipeline that encodes, renders, evaluates, and evolves image-producing individuals.
Build a Text-to-Image Generator (from Scratch)
Generative AI is a class of machine-learning models that creates new content resembling patterns learned from data.
- How it is used
- Creates novel content for text, image, audio, video, editing, design, education, medicine, and entertainment.
- How deep it goes
- Explains the definition, major architectures, diffusion process, applications, and originality debate without detailing full model implementation.
- Examples
- Uses ChatGPT, DALL-E 2, Imagen, Stable Diffusion, photo editing, medical diagrams, and generated essays or code.
- What you can do after
- Evaluate generative applications, explain diffusion-based image creation, and recognize originality, copyright, and deepfake risks.
Read how Build a Text-to-Image Generator (from Scratch) explains generative modeling →
Spring AI in Action
Generative AI is artificial intelligence that produces new content such as text, images, audio, or code from learned patterns.
- How it is used
- Powers Spring Boot applications that answer questions and generate content through natural-language interaction.
- How deep it goes
- Defines generated content, prompts, and nondeterminism, then demonstrates a simple REST application without detailing model internals.
- Examples
- Uses question answering, document summarization, sentiment analysis, content moderation, classification, image generation, and sky-color responses.
- What you can do after
- Build it from scratch as a Spring Boot question-and-answer service using Spring AI and a supported AI provider.
Just Use Postgres!
Generative AI comprises artificial-intelligence models that produce new content such as text, images, audio, or code from learned patterns.
- How it is used
- Builds recommendation services, chatbots, watchlist agents, and retrieval-augmented applications around generated content.
- How deep it goes
- Explains models, prompts, embeddings, tools, and application patterns without descending into model internals.
- Examples
- Uses movie recommendations, a Shrek filmography question, conversational chatbots, embeddings, and watchlist management.
- What you can do after
- Build it from scratch by combining LLMs, embeddings, Postgres, pgvector, similarity search, and RAG.
Related concepts
- language modelincludes · A language model is a computational model that represents patterns in linguistic tokens to estimate sequence probabilities and predict or generate language from context.
- Gaussian mixture modelincludes · A Gaussian mixture model is a probabilistic model representing a distribution as a weighted combination of multiple Gaussian component distributions.
- socio-technical systempart of · A socio-technical system is an interdependent arrangement of people, technology, processes, and social or organizational structures that interact to achieve outcomes.
- artificial intelligencekind of · Artificial intelligence is the field concerned with creating computational systems that perform tasks associated with human intelligence, such as learning, reasoning, perception, or decision-making.
- variational autoencoderincludes · A variational autoencoder is a generative neural network that learns a probabilistic latent representation of data and decodes samples from it to reconstruct or generate data.
- generative adversarial networkincludes · A generative adversarial network is a generative model with a generator and discriminator trained in competition, enabling the generator to produce samples resembling a target dataset or distribution.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.