Start free trial
Trending New Genres Authors
EnglishEnglish
EspañolSpanish
简体中文Chinese
繁體中文Chinese (Traditional)
FrançaisFrench
DeutschGerman
日本語Japanese
PortuguêsPortuguese
ItalianoItalian
한국어Korean
РусскийRussian
NederlandsDutch
العربيةArabic
PolskiPolish
हिन्दीHindi
Tiếng ViệtVietnamese
SvenskaSwedish
ΕλληνικάGreek
TürkçeTurkish
ไทยThai
ČeštinaCzech
RomânăRomanian
MagyarHungarian
УкраїнськаUkrainian
IndonesiaIndonesian
DanskDanish
SuomiFinnish
БългарскиBulgarian
עבריתHebrew
NorskNorwegian
HrvatskiCroatian
CatalàCatalan
SlovenčinaSlovak
LietuviųLithuanian
SlovenščinaSlovenian
СрпскиSerbian
EestiEstonian
LatviešuLatvian
فارسیPersian
മലയാളംMalayalam
தமிழ்Tamil
اردوUrdu
Searching...
SoBrief
Applied Artificial Intelligence

Applied Artificial Intelligence

An Introduction For Business Leaders
Buying a model is not buying an outcome: how leaders decide where AI belongs, and where it fails.
by Mariya Yao 2018 201 pages
3.57
410 ratings
Listen
Immersive
V2.3
Amazon Kindle Audible
Try Full Access for 3 Days
Unlock listening & more!
Continue

Key Takeaways

Most business AI is narrow automation, not a machine mind

Separate capability from hype. The book distinguishes today’s specialized AI from Artificial General Intelligence, or AGI, which would transfer learning across domains as people do. Deep Blue and AlphaGo could defeat human champions at chess and Go, but neither could use that skill to drive a car or write music. Impressive performance on one task is not general intelligence.

Recognize the working methods. The Machine Intelligence Continuum is a practical way to distinguish scripted systems, statistical predictions, and systems that learn or generate content. In supervised learning, labeled examples teach a model to predict categories or numbers. Unsupervised learning looks for patterns in unlabeled data. Reinforcement learning improves actions through feedback. These categories overlap, and simpler methods can outperform elaborate neural networks.

Analysis

This distinction redirects attention from dramatic speculation to the capabilities a business can actually buy, build, and evaluate. A system that recognizes fraud or ranks sales leads may be valuable without possessing anything like human understanding. The useful questions are whether it performs reliably, under which conditions, and how it fails. The continuum is a helpful map, not a permanent taxonomy: technologies blur categories, and labels can change faster than underlying business needs. Leaders should ask vendors to demonstrate the actual task and limits rather than accept an AI label as evidence of sophistication.

Begin with a costly, repeatable decision, then test whether AI fits

Start with the need. The strongest opportunities involve repeated work, measurable decisions, and enough relevant data to learn from. AI is a poor fit when the task requires subtle judgment, when reliable answers cannot be defined, or when ordinary software or human work is cheaper. The book’s practical test is whether the task can be broken into clear inputs, decisions, and outcomes, and whether useful data exists.

Design around real users. FarmGuide used satellite images to estimate Indian farms’ crop yields, helping lenders assess risk more fairly. UNICEF’s U-Report let young people report problems by SMS, a crucial choice where smartphones and reliable internet were scarce. In one Liberia poll, 86 percent of 13,000 respondents reported teachers exchanging grades for sex. The product’s accessibility helped surface a hidden problem.

Analysis

These examples show that successful applied AI begins with a human problem and a workable route to the people affected. That is closely related to service design, where the quality of the interaction can matter more than the sophistication of the back end. U-Report’s poll also illustrates a measurement caution: respondents who choose to answer are not necessarily representative of an entire population. The results can still expose a serious issue, but decision-makers should distinguish evidence of reported experience from a precise estimate of prevalence.

Audit the story your data tells before trusting its predictions

Data is constructed. People decide what to measure, how to collect it, and what counts as a correct result. The book calls observable, objective evidence that reflects reality ground truth. If labels are subjective, records are incomplete, or definitions are inconsistent, a model can learn the wrong lesson with impressive precision. Even a basic sales report depends on choices such as who qualifies as a customer, which quarter applies, and how returns are counted.

Errors enter at every stage. A website that always displays product A first may falsely suggest that customers prefer it. A survey available only online may miss older people without internet access. Sampling, capture, measurement, and processing can all distort what the data represents. More data does not automatically fix a biased or poorly defined dataset.

Analysis

The deeper point is epistemic: a model does not discover reality directly, but learns from a measurement system designed by people. That makes data documentation part of analytical quality, not clerical overhead. A useful audit asks who is represented, who is absent, and what behavior a recorded variable actually captures. Clicks, for instance, reveal interaction, not necessarily satisfaction or intent. The caution is not to reject quantitative evidence, but to state clearly what it can and cannot support.

Design for people missing from both the dataset and the room

Bias can hide in proxies. Removing race or gender from a model does not remove discrimination if other variables closely track those characteristics. The book cites cases in which online ads for high-paying jobs were less likely to reach women, and same-day delivery was unavailable in some predominantly Black neighborhoods. In healthcare, underrepresentation in digital records and clinical studies can leave models unreliable for groups such as pregnant women, older adults, and poorer communities.

Broaden responsibility. A diverse design team can notice overlooked needs, but inclusion is not a substitute for testing. The book also warns that AI can amplify malicious activity through deception, surveillance, cyberattacks, or autonomous weapons. Ethical design therefore requires accountable decision-making, transparency appropriate to the context, and repeated checks for harmful outcomes.

Analysis

Fairness is not a single technical setting. It involves decisions about the goal, the population, the cost of errors, and who has power to contest an outcome. This connects AI governance to procedural justice: people care not only whether a decision is accurate, but whether the process is understandable and open to correction. Diverse teams can expand the questions asked, although no team composition guarantees fair results. Audits, stakeholder participation, and remedies for people harmed by decisions remain essential.

Give AI a senior champion, shared infrastructure, and frontline allies

Prepare the organization first. AI projects struggle when departments hoard data, technology systems cannot communicate, or executives ignore analysis that conflicts with intuition. The book calls the highest-paid person’s opinion a HiPPO, a warning about decisions driven by rank rather than evidence. Shared data infrastructure, executive commitment, and a culture willing to test assumptions are prerequisites, not optional polish.

Build across departments. The proposed AI SWAT Team is a small, cross-functional group that identifies and prioritizes opportunities, coordinates work, and spots risks. The book suggests five to fifteen members, with technical staff alongside people from functions such as finance, HR, and operations. Frontline employees matter because they understand daily workflows and may resist automation if they are excluded or fear sudden job loss.

Analysis

This is a sociotechnical view of AI: outcomes depend on organizational incentives and relationships as much as on algorithms. A central team can set standards and prevent duplicated work, while local staff supply context that central experts lack. The tension is important. Excessive centralization can turn governance into a bottleneck; excessive decentralization creates incompatible systems and unclear accountability. A sound arrangement gives teams shared infrastructure and safeguards while preserving meaningful local input.

Buy or build only after ranking value, risk, and proof

Make the investment case. Compare an opportunity with business goals, likely benefits, full costs, risks, timelines, and stakeholder support. Decide whether the capability is central to the business, whether the organization has the data and talent to build it, and what maintenance will cost. Buying a vendor product can test an idea quickly; a company can later combine vendor and internal tools, or build a custom system when strategic value justifies it.

Choose a useful measure. A true north metric is a specific indicator that links everyday product behavior to a larger business goal. The book gives Facebook’s focus on users connecting with ten friends within seven days and Slack’s focus on teams exchanging 2,000 messages. Such measures are more informative than sign-ups alone, but they need to be checked against real customer and business outcomes.

Analysis

This approach treats AI investment as a portfolio of uncertain bets rather than a guaranteed upgrade. Staged pilots limit exposure and create evidence before a large commitment. The metric discussion also echoes Goodhart’s law: once a measure becomes a target, people may optimize the number while damaging the underlying goal. A useful scorecard therefore pairs a leading indicator with guardrails, such as customer experience, error rates, and the cost of serving users. A vendor’s performance metric should never be assumed to match the buyer’s.

Judge AI by its costly mistakes, not its accuracy headline

Count the errors that matter. Accuracy is the share of all predictions that are correct, but it can obscure whether a model misses the cases that matter most. Precision asks how often positive flags are correct. Recall asks how many of the true positive cases the model finds. In a spam filter, a false positive can hide an important message; in cancer screening, a false negative can miss a sick patient. The acceptable balance depends on the consequences.

Set thresholds deliberately. A model is not good merely because one summary number looks high. Leaders need to specify the relative costs of false alarms and missed cases, examine performance on relevant groups, and test on data not used to train the model. The book’s medical examples illustrate potential benefits, but reported performance figures depend on the study, data, and evaluation method.

Analysis

This is a decision-theory lesson: model quality is inseparable from the action taken after a prediction. A screening system that flags uncertain cases for a clinician has a different risk profile from one that automatically denies care. The same score can therefore be acceptable in one workflow and dangerous in another. Accuracy remains useful, but only alongside class balance, subgroup performance, and the consequences of each error. Human review helps only when reviewers have time, authority, and adequate information to challenge the model.

Treat deployment as maintenance, not a model's finish line

A model is one component. Data pipelines, software interfaces, security, deployment, monitoring, and updates often require more work than the model itself. Technical debt is the future cost created by quick fixes or fragile design choices. Machine learning systems can accumulate code, data, and model complexity that makes them expensive to understand and maintain.

Build a learning cycle. Agile development breaks work into short cycles, often two to four weeks, so teams can test value before committing heavily. The book describes PwC teams using small groups and repeated sprints to assess whether an idea deserves more investment. After launch, models need fresh data and performance checks. A recommendation system, for example, can become stale as products or customer behavior change. The book advises devoting substantial effort to measurement and upkeep, not just model creation.

Analysis

Deployment turns a prediction into an operating system that can affect customers and employees repeatedly. This makes monitoring analogous to safety engineering: teams need to detect drift, record changes, assign ownership, and know how to intervene when performance falls. The source’s emphasis on maintenance is especially valuable because development culture often rewards launches more visibly than reliability. A further discipline is to define rollback conditions in advance, so a degraded model can be paused without waiting for a crisis.

Automate tasks so people can spend more time on judgment

Target pieces of work. The book’s enterprise examples focus less on replacing entire departments than on relieving repetitive tasks. Finance tools can flag duplicate expenses; legal systems can locate contract clauses; HR tools can sort applications and answer routine policy questions. In customer support, a bot can handle basic questions while a human agent takes complex cases. Deloitte used language-processing tools to reduce one contract-review process from several months to about a week.

Preserve the human role. Marketing systems can test ad variations, sales models can prioritize likely leads, and analytics tools can prepare messy data. These tools can free people for relationship-building, complex judgment, and strategy. But automation only creates value when the freed time is put to productive use and the overall customer or employee experience improves.

Analysis

The task-level framing avoids treating occupations as indivisible units. It also reveals that augmentation is an organizational choice, not an automatic property of software. A firm can use saved time to improve service, increase workloads, or reduce staffing. Measuring only labor hours saved misses these differences. Evaluation should include service quality, employee discretion, and whether workers can actually shift into higher-value activities. Human oversight is most useful when it adds contextual judgment rather than merely rubber-stamping a machine output.

Budget for retraining and harm prevention alongside the software

Transitions need investment. Automation changes tasks and can eliminate jobs, while also creating demand for different skills. The book cites estimates that 45 percent of work tasks could be automated, while about 5 percent of jobs had been fully displaced at the time of the cited research. It also reports a sharp gap between executives’ automation plans and their training plans: 74 percent expected to automate, but only 3 percent expected to increase worker training.

Make adaptation concrete. Accenture used employee skill information to identify adjacent capabilities workers could develop. GE and BMW also invested in learning programs for their workforces. These examples support a practical obligation: involve employees, identify changing skill needs, and make training accessible before roles disappear. Leaders must also monitor effects on customers and the wider public, not just the company’s balance sheet.

Analysis

A technology transition distributes gains and costs unevenly. Training can help workers move into changing roles, but it cannot guarantee that every displaced employee will find an equivalent position; the quality and availability of new jobs matter too. This makes workforce planning part of responsible deployment, not a public-relations add-on. The same principle applies outside the firm: affected customers need ways to understand, challenge, and recover from consequential automated decisions. Productivity gains are more durable when the people bearing transition costs have a voice in shaping them.

Analysis

The book’s central insight is that enterprise AI is less an algorithm-selection problem than a coordination problem. An organization needs a worthwhile decision to improve, data that represents the relevant world, technical capacity to build and operate a system, and people empowered to act on its output. Each link can fail independently. A technically strong model trained on distorted records can scale an old mistake; a reliable prediction ignored by managers produces no value; a useful tool deployed without employee trust invites resistance.

The argument is strongest where it treats product design, organizational culture, and technical performance as connected. The SMS-based U-Report example shows that access and context may matter more than novelty. The book also correctly distinguishes model creation from the larger work of data preparation, integration, monitoring, and retraining. That perspective counters a common managerial illusion: the purchase of a model is not the purchase of an outcome.

Its limitations reflect both the subject and its 2018 publication date. Specific performance claims and vendor examples are snapshots, not timeless evidence. The broad estimate that AI can automate a share of tasks cannot predict who bears the resulting costs, and efficiency gains do not guarantee better jobs or shared prosperity. The book recognizes these risks, but its optimistic cases could be strengthened by treating labor bargaining power and the distribution of productivity gains as central design questions.

A useful modern reading is therefore neither technological enthusiasm nor blanket suspicion. Treat every AI initiative as a governed intervention in a workflow. Define the human objective, identify who may be excluded or harmed, compare against simpler alternatives, and establish how the system can be challenged or stopped. The goal is not to make an organization AI-first at any price. It is to make decisions and services better in ways that remain observable, accountable, and worth their social cost.

Summary in 30 Seconds
Most business AI is narrow automation, so start with a costly, repeatable decision backed by relevant data. Data is constructed by human choices, and bias hides in proxy variables, so audit it first. Judge models by the cost of false alarms versus missed cases, not headline accuracy. Deployment demands pipelines, monitoring, and retraining. Pair a senior sponsor with a cross-functional team, automate tasks rather than jobs, and fund worker retraining.
Contains spoilers

Last updated:

Report Issue

Review Summary

3.57 out of 5
Average of 410 ratings from Goodreads and Amazon.

Applied Artificial Intelligence receives mixed reviews, with an average rating of 3.58 out of 5. Readers appreciate its accessible overview of AI for business leaders and non-technical audiences. The book is praised for its clear explanations, practical advice, and structured approach to implementing AI initiatives. However, some critics find it too basic or lacking in-depth technical information. Many reviewers recommend it as a good starting point for understanding AI in business contexts, while others desire more detailed case studies and advanced content.

Your rating:
4.17
170 ratings
Want to read the full book?

Glossary

Artificial General Intelligence (AGI)

Cross-domain machine intelligence

A hypothetical form of machine intelligence able to learn from limited experience, form abstract concepts, and transfer knowledge between different domains at roughly human or greater capability. The book contrasts AGI with today’s narrow systems, which can perform specific tasks impressively but cannot readily apply that skill elsewhere.

Machine Intelligence Continuum

Capability map for AI

The authors’ framework for describing machine systems according to the kinds and complexity of work they perform, from following fixed rules to predicting, learning, creating, relating, and more speculative forms of intelligence. Its categories overlap rather than form strict technical boundaries. It is intended to help business leaders distinguish practical capabilities from inflated claims.

HiPPO

Highest-paid person’s opinion

A decision-making pattern in which the view of the most senior or highly paid person overrides evidence and analysis. The term highlights how hierarchy can derail data-informed work, even when leaders believe they are relying on experience or vision. The authors use it to explain why analytical culture and executive support both matter.

ground truth

Verified reference evidence

The observable, objective evidence used as a reference for training or evaluating a model. The term does not mean that a dataset is automatically true: its reliability depends on how the phenomenon was defined, measured, labeled, and recorded. If the reference labels are wrong or incomplete, a model’s apparent performance can be misleading.

AI Strategy Framework

Opportunity-ranking method

The authors’ method for comparing potential AI projects against business goals and practical constraints. It prompts leaders to consider strategic fit, opportunity size, investment, expected return, risk, timeline, and stakeholder support. Its purpose is to prioritize viable business problems rather than select technology first and search afterward for a use.

true north metric

Core progress indicator

A clear, actionable measure chosen to represent whether a project or product is moving toward a meaningful business outcome. It should be more informative than a vanity statistic, understandable across teams, and aligned with customer and organizational goals. The book emphasizes that teams should verify the measure’s data quality and avoid changing it casually.

AI SWAT Team

Cross-functional AI group

A small group of technical staff, business leaders, and representatives from different functions and levels of an organization. The team identifies and ranks automation opportunities, coordinates implementation, communicates the purpose of projects, and surfaces organizational or ethical risks. The book suggests a typical size of five to fifteen members.

technical debt

Future cost of shortcuts

The additional work and expense created when an early design shortcut makes software harder to change, operate, or maintain. In machine learning, debt can involve code, training data, or model complexity, along with dependencies across data pipelines and production systems. A model may work at first while becoming increasingly costly to monitor and update.

Machine Learning as a Service (MLaaS)

Hosted machine-learning tools

A platform or vendor service that supplies some or all of the infrastructure and tools needed to develop, train, deploy, or operate machine-learning models. It can help organizations lacking specialized staff or systems adopt capabilities more quickly. Buyers still need to assess data access, security, integration, cost, performance, and ongoing support.

FAQ

1. What is "Applied Artificial Intelligence: An Introduction For Business Leaders" by Mariya Yao about?

  • Practical AI Guide for Leaders: The book is a practical handbook designed to help business leaders understand, plan, and implement artificial intelligence (AI) initiatives in their organizations.
  • Bridges Technical and Business Worlds: It balances technical explanations with actionable business strategies, avoiding both overly technical jargon and vague generalizations.
  • Focus on Enterprise Impact: The content is tailored to enterprise-scale challenges, emphasizing how AI can enhance productivity, drive innovation, and solve real-world business problems.
  • Ethics and Societal Impact: The book also addresses the ethical, social, and workforce implications of AI, advocating for responsible and inclusive AI development.

2. Why should I read "Applied Artificial Intelligence" by Mariya Yao?

  • Executive-Focused Content: The book is specifically written for business leaders, executives, and entrepreneurs who want to leverage AI without needing a deep technical background.
  • Actionable Frameworks: It provides step-by-step frameworks for building AI-ready cultures, recruiting talent, and planning successful AI projects.
  • Real-World Case Studies: The book includes numerous examples and case studies from various industries, illustrating both the promises and pitfalls of AI adoption.
  • Ethical Guidance: It emphasizes the importance of ethical AI, diversity, and inclusion, helping leaders avoid common mistakes and unintended consequences.

3. What are the key takeaways from "Applied Artificial Intelligence" by Mariya Yao?

  • AI is Not Magic: AI is a set of evolving technologies, not a silver bullet; understanding its capabilities and limitations is crucial for success.
  • Data and Culture Matter: Building a data-driven, AI-ready culture and having clean, relevant data are foundational to any successful AI initiative.
  • Strategic Implementation is Key: Leaders must prioritize business goals, perform opportunity analysis, and use structured frameworks to identify high-ROI AI projects.
  • Ethics and Inclusion: Proactively addressing bias, ethics, and workforce impact is essential to building responsible and sustainable AI solutions.

4. What are the best quotes from "Applied Artificial Intelligence" and what do they mean?

  • “AI is like teenage sex. Everyone talks about it. Nobody knows how to do it. But everyone thinks everyone else is doing it, so they claim to do it too.” — Highlights the hype and confusion surrounding AI adoption in business.
  • “Data is not reality.” — Reminds readers that data is a human construct, subject to bias and error, and must be carefully managed for AI to be effective.
  • “Tools are not meant to make our lives easier. They are meant to give us leverage so that we can push harder. Tools lift rocks. People build cathedrals.” — Emphasizes that AI should empower human creativity and strategic thinking, not replace it.
  • “The future is already here—it’s just not evenly distributed.” — Points to the uneven adoption and impact of AI across industries and societies.

5. How does "Applied Artificial Intelligence" by Mariya Yao define and differentiate key AI concepts like AI, AGI, machine learning, and deep learning?

  • AI vs. AGI: AI refers to systems that perform specific tasks (narrow or weak AI), while AGI (Artificial General Intelligence) is human-level intelligence capable of abstract reasoning and knowledge transfer—something we have not yet achieved.
  • Machine Learning: Machine learning is a subset of AI that enables computers to learn from data without explicit programming, using techniques like supervised, unsupervised, and reinforcement learning.
  • Deep Learning: Deep learning is a further subset of machine learning, using multi-layered neural networks to achieve human-level performance in tasks like image and speech recognition.
  • Other Approaches: The book also discusses symbolic systems, probabilistic programming, and ensemble methods, highlighting their roles and limitations in enterprise AI.

6. What is the Machine Intelligence Continuum in "Applied Artificial Intelligence" by Mariya Yao, and why is it important?

  • Seven Levels of Intelligence: The Machine Intelligence Continuum (MIC) categorizes AI systems into seven levels: Systems That Act, Predict, Learn, Create, Relate, Master, and Evolve.
  • Functional Differentiation: Each level represents increasing complexity and capability, from simple rule-based automation to hypothetical superintelligent, self-evolving systems.
  • Practical Relevance: Most current enterprise AI solutions fall into the "Act," "Predict," and "Learn" categories, helping leaders set realistic expectations.
  • Strategic Planning Tool: Understanding the continuum helps executives evaluate vendor claims, plan AI investments, and anticipate future developments.

7. What are the main promises and challenges of AI according to "Applied Artificial Intelligence" by Mariya Yao?

  • Promises: AI can drive social good (e.g., microfinance, social justice, medical diagnosis), improve efficiency, and unlock new business opportunities.
  • Challenges: Risks include biased or incomplete data, lack of diversity in development teams, malicious uses of AI, and unintended societal consequences.
  • Ethical Imperative: The book stresses the need for ethical governance, inclusive design, and democratized access to AI education and resources.
  • Real-World Examples: Case studies illustrate both the transformative potential and the dangers of poorly implemented or unethical AI.

8. How does "Applied Artificial Intelligence" by Mariya Yao recommend business leaders develop an enterprise AI strategy?

  • Build an AI-Ready Culture: Assess organizational readiness, secure executive and board-level buy-in, and foster a data-driven mindset.
  • Form a Multi-Disciplinary Team: Assemble an "AI SWAT Team" with cross-functional expertise to identify, prioritize, and execute AI projects.
  • Strategic Frameworks: Use gap analysis, SWOT, and the book’s AI Strategy Framework to evaluate opportunities and risks.
  • Pilot and Scale: Start with small, high-ROI projects, demonstrate early wins, and scale successful initiatives across the organization.

9. What advice does "Applied Artificial Intelligence" by Mariya Yao give on building and recruiting technical AI talent?

  • Understand Roles: Clarifies distinctions between data scientists, machine learning engineers, researchers, and data engineers.
  • Seek Key Qualities: Values mathematical aptitude, curiosity, creativity, perseverance, rapid learning, and passion for the business problem.
  • Recruiting Strategies: Suggests university partnerships, hackathons, specialized training programs, and retraining existing staff as alternatives to traditional hiring.
  • Emphasize Unique Advantages: Attract talent by offering access to quality data, diverse problems, strong teams, and meaningful impact.

10. How does "Applied Artificial Intelligence" by Mariya Yao guide organizations in planning, implementing, and measuring AI projects?

  • Define Clear Goals: Prioritize business objectives and select specific, actionable "true north" metrics for each project.
  • Data and Analytics Readiness: Ensure data quality, centralization, and analytics maturity before launching AI initiatives.
  • Build vs. Buy Decision: Weigh the pros and cons of developing in-house solutions versus leveraging third-party vendors, considering data, talent, cost, and strategic fit.
  • ROI and Portfolio Approach: Calculate expected returns, allocate budgets, and manage AI projects as a portfolio of experiments with varying risk and reward.

11. What are the common pitfalls in data collection, preparation, and machine learning model development highlighted in "Applied Artificial Intelligence" by Mariya Yao?

  • Data is Not Reality: Data can be biased, incomplete, or misinterpreted; ground truth must be carefully established.
  • Common Data Errors: Includes undefined goals, definition, capture, measurement, processing, coverage, sampling, inference, and unknown errors.
  • Model Mistakes: Underfitting, overfitting, and lack of generalization are frequent issues; rigorous validation and continuous monitoring are essential.
  • Iterative Process: Emphasizes the need for agile development, ongoing experimentation, and maintenance to manage technical debt and ensure long-term model performance.

12. How does "Applied Artificial Intelligence" by Mariya Yao address the ethical, workforce, and societal impacts of enterprise AI?

  • Ethical Responsibility: Leaders must proactively address bias, discrimination, and the potential for malicious use of AI.
  • Workforce Transformation: AI will automate routine tasks but also create demand for new skills; continuous training and upskilling are vital.
  • Inclusive Design: Advocates for diverse, multidisciplinary teams and collaborative design to ensure AI benefits all stakeholders.
  • Long-Term Vigilance: Encourages ongoing assessment of AI’s impact on employees, customers, and society, with a commitment to do no harm and uphold human values.

About the Author

Mariya Yao is an author and expert in the field of artificial intelligence. She co-wrote "Applied Artificial Intelligence" to provide business leaders and non-technical readers with a comprehensive guide to implementing AI in corporate environments. Yao's approach focuses on demystifying AI concepts and offering practical strategies for organizations to leverage AI technologies. Her work emphasizes the importance of understanding AI's business applications, ethical considerations, and potential impact on various industries. Yao's expertise in explaining complex AI topics in accessible language has made her book a valuable resource for those seeking to navigate the rapidly evolving landscape of artificial intelligence in business.

Download PDF

To save this Applied Artificial Intelligence summary for later, download the free PDF. You can print it out, or read offline at your convenience.
Download PDF
File size: 0.41 MB     Pages: 22

Download EPUB

To read this Applied Artificial Intelligence summary on your e-reader device or app, download the free EPUB. The .epub digital book format is ideal for reading ebooks on phones, tablets, and e-readers.
Download EPUB
File size: 1.38 MB     Pages: 29
Want to read the full book?
Follow
Listen1 mins
Home
Swipe
Library
Get App
Try Full Access for 3 Days
Listen, bookmark, and more
Compare Features Free Pro
📖 Read Summaries
Read unlimited summaries. Free users get 3 per month
🎧 Listen to Summaries
Listen to unlimited summaries in 40 languages
—
❤️ Unlimited Bookmarks
Free users are limited to 4
—
📜 Unlimited History
Free users are limited to 4
—
📥 Unlimited Downloads
Free users are limited to 1
—
Risk-Free Timeline
Today: Get Instant Access
Listen to full summaries of 26,000+ books. That's 12,000+ hours of audio!
Day 2: Trial Reminder
We'll send you a notification that your trial is ending soon.
Day 3: Your subscription begins
You'll be charged on Nov 24,
cancel anytime before.
Consume 2.8× More Books
2.8× more books Listening Reading
Our users love us
1M+ readers
Trustpilot Rating
TrustPilot
4.6 Excellent
This site is a total game-changer. I've been flying through book summaries like never before. Highly, highly recommend.
— Dave G
Worth my money and time, and really well made. I've never seen this quality of summaries on other websites. Very helpful!
— Em
Highly recommended!! Fantastic service. Perfect for those that want a little more than a teaser but not all the intricate details of a full audio book.
— Greg M
Save 62%
Yearly
$119.88 $44.99/year/yr
$3.75/mo
Monthly
$9.99/mo
Start a 3-Day Free Trial
3 days free, then $44.99/year. Cancel anytime.
Unlock a world of fiction & nonfiction books
26,000+ books for the price of 2 books
Read any book in 10 minutes
Discover new books like Tinder
Request any book if it's not summarized
Read more books than anyone you know
#1 app for book lovers
Lifelike & immersive summaries
30-day money-back guarantee
Download summaries in EPUBs or PDFs
Cancel anytime in a few clicks

Why one summary?

Scanner
Find a barcode to scan

⭐
⭐
⭐
We have a special gift for you
Open
38% off
DISCOUNT FOR YOU
⭐
⭐
⭐
$79.99
$49.99/year
Only $4.16 per month
Continue
2 taps to start, super easy to cancel
Settings
General
Widget
Loading...