Chapter 1: Artificial intelligence in IT

Introduction

This introduction frames generative AI as a practical, transformative aid for IT work that must be paired with human judgment and used for real workflow improvements.

• Generative AI impact on IT: Explain how generative AI is changing IT work by accelerating routine tasks and becoming important knowledge for IT professionals.
  Why it matters: This establishes why AI skills are becoming part of effective day-to-day IT practice.
• IT workflow automation examples: Describe the examples of AI improving IT workflows: extracting, checking, formatting, and inserting PDF data into a database, and detecting unusual security-related patterns in logs or user behavior.
  Why it matters: These examples show that AI can remove repetitive work and help teams respond to potential problems sooner.
• Human judgment alongside AI: Explain why AI should complement human expertise rather than replace it: AI provides speed, while people supply empathy, creativity, deep contextual understanding, and judgment.
  Why it matters: Using AI effectively requires people to evaluate its role and apply professional judgment to its outputs.

Key terms: generative AI, AI tools, IT industry, PDFs, databases, application logs, user behavior, security problems, human expertise, human judgment

1.1 Generative AI changes everything

This section introduces improving generative AI capabilities and shows concrete ways AI assistance can support everyday IT and workplace tasks.

• Improving AI reasoning: Describe how the comparison between ChatGPT-3.5 and GPT-4 illustrates improving AI reasoning, using the stacking task as an example. State that newer models also have stronger analysis and image-processing capabilities, while AI models still have drawbacks.
  Why it matters: Recognizing both the progress and limitations of AI helps you use its outputs appropriately at work.
• Natural-language AI assistance: Explain how generative AI enables a conversational way of working: a person can ask a chatbot about technical problems in natural language, much as they would ask a colleague. Describe this as assistance for practical work rather than a guarantee of correct answers.
  Why it matters: Conversational interaction makes AI tools accessible for IT tasks without requiring prior AI expertise.
• Everyday IT use cases: Identify the practical IT and workplace uses presented for AI assistance, including analyzing logs for root causes, extracting insights from data, drafting disaster recovery plans, writing reviews or professional emails, and creating scripts. Explain that the book presents these as real daily uses rather than merely theoretical possibilities.
  Why it matters: These examples help you recognize work tasks where AI can save effort and support better outcomes.
• Experiment-driven opportunity discovery: Explain that experimenting with AI revealed expanding uses, such as documentation, evaluations, intelligent data extraction, and batch renames. State that unfamiliar use cases can become visible when people try the tools on practical tasks.
  Why it matters: Trying AI on real work can uncover useful applications that may not be obvious at first.

Key terms: generative AI, ChatGPT-3.5, GPT-4, reasoning, image-processing capabilities, AI chatbot, ChatGPT, queries, prompts, natural conversations

1.2 The elephant in the server room

This section argues that AI will not replace skilled IT professionals but will make AI-capable professionals more valuable by amplifying their existing expertise.

• AI and IT employment: Explain the section's view that AI is not coming to replace IT workers outright, although it may reduce some open positions while also creating new opportunities. State why effective AI use is expected to become a job requirement and increase an IT professional's value.
  Why it matters: The career advantage comes from learning to use AI effectively rather than treating it as a substitute for IT expertise.
• Expertise amplification by AI: Describe why the section considers AI most effective in the hands of skilled IT professionals with domain knowledge. Explain that users must ask appropriate questions and implement AI's answers correctly, so AI amplifies existing skills rather than reliably disguising incompetence.
  Why it matters: AI output is useful only when a professional can frame the task and judge and apply the result appropriately.
• Limits of AI substitution: State the evidence and conclusion from the accounting-friend example: a non-IT professional could not progress beyond the first few steps of an IT task with ChatGPT, illustrating that current AI is not a magic wand for nonprofessionals.
  Why it matters: This example reinforces why AI assistance does not remove the need for genuine IT competence, especially in consequential work.

Key terms: AI, open positions, new opportunities, AI will not replace you, but the person using AI will, domain, skilled professionals, ask the right questions, implement the answers appropriately, amplifies existing skill sets, non-IT pro

1.3 Strengthening your skill set

This section shows how AI expands IT professionals' capabilities by automating and documenting work, supporting unfamiliar planning tasks, and helping teams communicate more constructively.

• AI-assisted task orchestration: Describe how AI can handle the planning, coding, documentation, and follow-up communications for IT work, shifting an IT professional's role toward orchestrating systems rather than performing every task manually.
  Why it matters: This lets you use AI to reduce routine work while retaining responsibility for directing the overall outcome.
• Expanded automation opportunity: Explain why AI makes automation worth considering more often: it can create code quickly enough that the time and effort required to build an automation are much lower than with manual development.
  Why it matters: A lower effort to automate means you can look for efficiency opportunities that previously would not have justified custom development.
• AI-supported professional growth: Explain how AI frees people from routine tasks for complex problem-solving and future planning, while also helping an engineer perform unfamiliar project-management work such as selecting an approach and documenting it clearly.
  Why it matters: AI can help you broaden the kinds of work you can contribute to, rather than merely completing existing tasks faster.
• Blameless after-action reporting: Trace how ChatGPT helped rewrite a blame-focused after-action report into a comprehensive, unbiased account that emphasized systems, processes, communication gaps, and safeguards rather than attacking an individual.
  Why it matters: Factual, process-focused incident reporting can support problem-solving without worsening team conflict.

Key terms: automation, orchestrating systems, planning, coding, documenting, automate, code quickly, complex problems, planning for the future, project management

1.4 An introduction to generative AI

This section defines generative AI as data-trained systems that create new content and introduces the text and image generators used throughout the book.

• Generative AI definition: Explain that generative AI systems learn patterns from existing data and apply those patterns to create new content.
  Why it matters: This definition helps you recognize what makes generative AI useful for creating work outputs rather than only analyzing information.
• Text and image generators: Describe the two generative-AI types emphasized in the book: text generators, also called large language models, and image generators, including the kinds of content each can create or modify.
  Why it matters: Knowing which kind of generator fits a task lets you choose AI tools appropriately for written or visual work.

Key terms: generative AI, existing data, pattern learners, new content, text generators, large language models, LLMs, image generators, OpenAI GPT, Google Gemini

1.4.1 Language model chatbots

This section introduces language model chatbots, positions ChatGPT among them, and compares leading alternatives by their distinctive strengths and integrations.

• ChatGPT category membership: Explain why ChatGPT can simultaneously be described as an AI, a language model, a generative AI system, and a chatbot.
  Why it matters: These overlapping categories provide the vocabulary used to discuss AI tools throughout the book.
• Leading chatbot landscape: Describe the section's view of the leading chatbot market: ChatGPT changed perceptions of AI chatbots after its late-2022 launch, but competitors have caught up or sometimes moved ahead with different strengths.
  Why it matters: Choosing an AI assistant means comparing current capabilities rather than assuming the best-known chatbot is always best.
• Assistant-specific strengths: Distinguish the strengths attributed to Claude, Gemini, and Microsoft Copilot for Microsoft 365, including Claude's large-text handling, Gemini's Google Workspace-connected search, and Copilot's Microsoft business-tool integration and extra licensing requirement.
  Why it matters: Matching a chatbot to its ecosystem or task can make AI assistance more useful in day-to-day IT work.
• Availability and quality caveats: State that AI features may be released first in the United States and UK and that the authors selected both ChatGPT and Claude because, in their experience, they produced the highest-quality and most accurate responses for different purposes.
  Why it matters: Availability can vary by region, and tool choices should be based on practical fit and evaluated output quality.

Key terms: AI, language model, generative AI system, chatbot, ChatGPT, OpenAI, Anthropic Claude, Google Gemini, Microsoft Copilot, Claude

1.4.2 Image generation with AI models

This section contrasts artistic image generation with code- and markup-based AI outputs that give IT professionals more control over technical visuals and documentation.

• Creative versus technical visuals: Distinguish creative AI-generated images from technical visual work, including why artistic outputs are often less suitable when an IT task requires repeatable or precise results.
  Why it matters: Choosing the right kind of AI output prevents attractive but unreliable images from being used where precision is required.
• Markup-based technical diagrams: Describe how AI can generate Mermaid or PlantUML markup for flowcharts, sequence diagrams, and architecture diagrams, and explain why this is useful for documentation and presentations.
  Why it matters: Markup gives IT professionals a practical way to turn AI assistance into diagrams that can be used in technical communication.
• Controlled code-based mockups: Explain how AI-generated LaTeX and HTML/CSS can support technical documentation and mockups, especially because code can preserve exact text, layouts, and user control.
  Why it matters: Generating code rather than relying only on artistic images can make technical visuals more accurate and easier to control.
• Diagram format limitations: Explain that AI can design and describe diagrams well but does not directly create Microsoft Visio files, so extra tools or steps may be needed to reach enterprise formats.
  Why it matters: Knowing the format boundary helps you plan a realistic workflow instead of expecting AI output to be ready for every enterprise tool.

Key terms: creative images, technical images, AI art, DALL-E, Stable Diffusion, Midjourney, Mermaid, PlantUML, flowcharts, sequence diagrams

1.5 Service costs

This section explains how free tiers, paid subscriptions, and API playgrounds affect the cost of using AI services for the book's examples.

• Free-tier access and limits: Describe how free tiers for chatbots such as ChatGPT and Claude let readers use key models and features for learning, while imposing usage limits.
  Why it matters: You can begin practicing AI techniques without assuming that a paid subscription is required.
• Subscription value decision: Explain when a paid AI-service subscription may be worthwhile: when a user regularly reaches free-tier usage limits or needs extras such as faster processing and advanced features.
  Why it matters: This helps you choose paid access based on actual work needs rather than assuming everyone needs the same plan.
• API playground alternative: Describe vendor API playgrounds as chatbot-like web interfaces that can offer a lower-cost alternative to subscriptions, with cost depending on token usage.
  Why it matters: A playground can be a practical option when a chatbot subscription costs more than your actual level of use justifies.
• Current service pricing reference: State that Table 1.5 presents current pricing for popular AI services as of February 2025.
  Why it matters: Pricing changes over time, so the table should be treated as a dated reference rather than a permanent cost guarantee.

Key terms: free tier, usage limits, ChatGPT, Claude, paid plans, paid subscription, faster processing, advanced features, vendor API playgrounds, tokens

1.6 Using AI responsibly at work

This section frames the book as practical implementation guidance for IT professionals in organizations that have approved AI use, rather than a debate about broad AI ethics.

• Book’s practical AI scope: Explain that the book is for IT professionals whose organizations have approved AI use and who need guidance on implementing it effectively in IT work.
  Why it matters: This establishes that the book focuses on using approved AI tools productively in day-to-day IT work.
• Practical rather than ethical debate: Distinguish the section’s practical focus from a broad debate about AI ethics, including privacy, security, and plagiarism.
  Why it matters: Knowing this boundary helps readers understand why the book emphasizes actionable IT guidance rather than general ethical argument.

Key terms: approved AI use, implementation guidance, IT work, broad AI ethics, privacy, security, plagiarism, practical guidelines

1.6.1 Data security with AI

This section explains the data-security risks of sending company information to cloud AI services and identifies safer enterprise AI options.

• Cloud AI data exposure: Explain why using an AI service can expose company data outside the company network and what precautions are required before sharing confidential or protected information.
  Why it matters: Knowing where AI-bound data goes helps you avoid exposing protected company information.
• Enterprise AI security controls: Describe the enterprise AI options named in the section and how existing security settings, compliance rules, private endpoints, and IAM controls can make organizational AI use safer.
  Why it matters: Enterprise controls can let teams use AI while aligning it with their existing security and compliance requirements.

Key terms: cloud processing, confidential information, GDPR, protected data, private subscription, self-hosted instance, enterprise AI plans, Microsoft 365 Copilot, Azure OpenAI services, AWS Bedrock

1.6.2 Bias and fairness

This subsection explains how AI can reproduce bias in hiring and how people can detect and reduce unfair outcomes.

• Training-data bias in hiring: Explain how an AI system can reflect biases in its training data and produce unfair hiring recommendations. Describe the Amazon example in which a tool favored candidates named Jared who played lacrosse because those traits appeared among past successful hires.
  Why it matters: Recognizing that AI can inherit historical patterns helps you avoid treating its hiring output as objective.
• Fair AI hiring safeguards: Describe the practical safeguards for using AI fairly in hiring: ask every candidate the same questions, remove names from resumes, use skill-based tests, include diverse interviewers, and check decisions for bias patterns.
  Why it matters: These checks preserve AI's speed while reducing the chance that it unfairly screens or evaluates candidates.
• Bias checks on AI outputs: Describe how people can review AI-generated language and decisions for bias, using tools such as Textio to flag biased job-posting language and correcting defaults such as male pronouns for engineers. Explain why a diverse team helps find and fix such problems early.
  Why it matters: Reviewing outputs and involving diverse people catches biased language or assumptions that automated tools may otherwise spread.

Key terms: biases, training data, resume screening, hiring decisions, same interview questions, remove names from resumes, skill-based tests, diverse interviewers, bias patterns, Textio

1.6.3 AI writing and plagiarism

This section explains why AI-written text cannot be reliably detected and how source citations and disclosure support responsible AI-assisted writing.

• Limits of AI detection: Explain why AI-writing detectors cannot reliably determine whether content was written by AI, including the example that OpenAI discontinued its own detector because it was not accurate enough.
  Why it matters: You should not treat a detector result as proof that writing is or is not AI-generated.
• Citation features across AI tools: Distinguish AI tools that use web search and provide citations from tools that rely only on training data, whose outputs cannot be checked against outside sources. Describe Microsoft Copilot's automatic source citations.
  Why it matters: Citations let you check claims and give proper credit instead of treating an AI response as an unsupported authority.
• Transparent AI authorship: Describe responsible AI-assisted writing as being open about AI use and properly crediting the underlying sources, rather than relying on detection tools.
  Why it matters: Transparency and source credit help preserve credibility when AI contributes to written work.

Key terms: AI writing detectors, AI-written content, OpenAI detector, web search, search results, source citations, training data, Microsoft Copilot, ChatGPT, responsible AI use

1.6.4 Citing prompts in your work

This section explains how to disclose and cite AI use appropriately, use chatbots as assistance rather than authorship, and recognize the limits of AI-generated writing.

• AI use citation practices: Explain why AI-generated content is treated as nonrecoverable and how to disclose AI use by mentioning it in the text and citing the AI software according to applicable rules.
  Why it matters: Clear disclosure lets readers understand when AI contributed to work whose chatbot outputs cannot be verified like ordinary sources.
• AI-assisted work boundaries: Describe ways AI can assist work without writing the content itself, including brainstorming, proofreading, rubric checks, practice grading, and writing improvement.
  Why it matters: Using AI as a helper can improve work while preserving your own judgment and authorship.
• Limits of AI-authored books: Explain the chapter-writing experiment’s conclusion that an AI-generated outline may provide a useful idea, but AI alone did not produce a high-quality, understandable, useful book.
  Why it matters: Recognizing polished but empty AI prose helps you retain critical judgment over AI-generated material.

Key terms: nonrecoverable, AI-generated content, citation, OpenAI, ChatGPT, APA, brainstorm, proofread, rubrics, practice assignments

1.7 How we are using AI

This section explains how the authors use AI as a guided editorial collaborator, while retaining human judgment and using image generation as a practical exception.

• AI editorial collaboration: Describe how the authors use AI as an editorial assistant for drafts and ideas, including asking it to improve clarity and to review completed sections for accuracy, balance, length, and technical detail.
  Why it matters: Treating AI as a review and revision partner can improve professional content without handing over authorship.
• Guided professional prompting: Explain that when requesting writing help, the authors give AI explicit guidance about the desired tone, words to avoid, key points to include, and the need for a clear, direct message.
  Why it matters: Specific instructions help turn an AI response into content suited to a professional purpose.
• Human oversight and image generation: Distinguish the authors' view of AI as a collaborator rather than an expert from their use of DALL-E 3 for images, and state why human guidance remains necessary for professional writing.
  Why it matters: Knowing AI's limits helps you use it productively without mistaking generated output for expert human reasoning.

Key terms: editorial assistant, rough drafts, accuracy, clarity, technical detail, tone, words to avoid, key points, clear, directly

1.8 Prompts used in this chapter

This section shares the concrete AI prompts used to develop the chapter and positions such prompt lists as reusable inspiration for readers' work.

• Chapter prompt examples: Describe the kinds of concrete prompts the authors used with paid ChatGPT and Claude subscriptions while writing this chapter. Explain that the examples include generating practical IT-use ideas, revising wording and tone, restructuring prose, changing narrative perspective, improving transitions, and reviewing a revised chapter.
  Why it matters: These examples show that AI can support many small, specific writing and editing tasks rather than only generating an entire document.
• Reusable prompt-list purpose: Explain why the book includes a prompt list at the end of this and all later chapters: to give readers ideas for applying similar AI requests in their own work. State that these are concrete examples of how the authors used AI, not a claim that readers must use identical prompts.
  Why it matters: Seeing specific examples can help you adapt AI use patterns to your own tasks and workplace needs.

Key terms: ChatGPT, Claude, paid subscriptions, prompts, prompts used in this chapter, ideas for how to use AI in your own work

1.9 Summary

This summary presents AI as a practical accelerator for IT work that requires responsible use and continued human expertise.