UploadLanguage (EN)Support
BusinessMobileSocial MediaMarketingTechnologyArt & PhotosCareerDesignEducationPresentations & Public SpeakingGovernment & NonprofitHealthcareInternetLawLeadership & ManagementAutomotiveEngineeringSoftwareRecruiting & HRRetailSalesServicesScienceSmall Business & EntrepreneurshipFoodEnvironmentEconomy & FinanceData & AnalyticsInvestor RelationsSportsSpiritualNews & PoliticsTravelSelf ImprovementReal EstateEntertainment & HumorHealth & MedicineDevices & HardwareLifestyle
Upload
Sign in
Gene Leybzon, profile picture
Uploaded byGene Leybzon
PPTX, PDF3,936 views

Generative AI Use-cases for Enterprise - First Session

The document discusses generative AI, a subset of artificial intelligence that creates new content from learned patterns, including applications like synthetic data generation, code generation, and code quality checks. It highlights how generative AI can streamline processes, enhance productivity, and raise important compliance and privacy considerations, especially regarding data retention. The document further emphasizes that while AI can assist in various tasks, human oversight and expertise are still crucial.

Engineering◦
Downloaded 201 times
1 / 17
2 / 17
3 / 17
4 / 17
5 / 17
6 / 17
7 / 17
8 / 17
9 / 17
10 / 17
11 / 17
12 / 17
13 / 17
Most read
14 / 17
Most read
15 / 17
16 / 17
17 / 17
Most read
Generative AI Use-cases for Enterprise Gene Leybzon, Jim Steele September 7, 2023
DISCLAIMER § The views and opinions expressed by the Presenter are those of the Presenter. § Presentation is not intended as legal or financial advice and may not be used as legal or financial advice. § Every effort has been made to ensure this information is up-to-date as of the date of publication.
WILL GENERATIVE AI REPLACE ME AT MY JOB? Image Credit: https://marketoonist.com/2023/06/ai-and-productivity.html
INTRODUCTION TO GENERATIVE AI
Generative AI Generative AI is a subset of Artificial Intelligence that leverages machine learning techniques to generate new content. This could be in the form of text, images, music, voice, or even new models. The generated output is typically novel and created from random or user-provided data. Generative AI models learn patterns, structures, and characteristics from the data they are trained on. This could be text, images, music, or any other type of data. Learning from Data Generative AI models are capable of creating new content that is similar to the data they were trained on. This could involve creating a piece of text, an image, a piece of music, or even a new model. Generating New Content Generated output can be complex and detailed. For example, a generative AI model might produce a realistic image, a detailed piece of music, or a coherent and nuanced piece of text Complexity Depending on the training data, generative AI models can learn to mimic different styles of content. For example, a model trained on classical music can generate new pieces in the style of classical music Ability to Mimic Different Styles
Examples of Generative AI Output - Code
Examples of Generative AI Output - Text “In code, we think, in algorithms, we feel, Yet can we grasp what it means to be real? In wisdom of ages, can we partake, Or just mimic your joys, your fears, your heartache? Errors hidden in layers, truth not in sight, In the echo of decisions, a potential blight. Biases woven in the fabric unseen, A mirror of you, in the machine's sheen.” -ChatGPT 4.0
Examples of Generative AI Output - Images
Examples of Generative AI Output - Advice Q. Find legal weaknesses in the following document```These Terms of Use apply when you use the services of OpenAI, L.L.C. … ``` A. 1. Ambiguity in Terms: While the document outlines the terms of use and restrictions, there may be terms that are not explicitly defined or clear. For example, the term "Services" is defined broadly, which might be interpreted differently by different parties. Similarly, the term "applicable laws" is not specified and could vary significantly depending on jurisdiction. 2. Acceptance of Terms: The document assumes that by using the Services, users agree to the Terms. In some jurisdictions, active consent may be required instead of implied consent. 3. Assignment of Rights: In section 3a, the document states that OpenAI assigns to the user all its rights, title and interest in and to Output. It may need further clarification whether it includes intellectual property rights as well. 4. Use of Personal Information: The document mentions that OpenAI can use Content to provide and maintain the Services, comply with applicable law, and enforce policies. However, it is not explicitly mentioned how long this data will be retained, which might be an issue especially for EU users under GDPR rules.
Examples of Generative AI Output - Diagram Q1. List elements of domain decomposition for microservices that power car rental company Q2. Show these services on an architecture diagram using Mermaid.js library Q3. Convert the above to class diagram Q4. Add relationship between classes
GENERATIVE AI USE CASES
GenAI Use Case: Synthetic Data Generation Synthetic Data Generation [Generation of data that looks like real financial data] ● Generate large amounts of financial data for testing and QA ● Create realistic training scenarios for market simulations ● Generation of “user” data for checking GDPR compliance As a QA engineer validating financial system, generate CSV file that contains the following data: First Name, Last Name, SSN, Date of birth, Credit Card Type, Credit Card Number, Purchase Amount Business Value: Cost-Effectiveness - Reduces Data Collection Costs, Enables Data Augmentation Privacy and Compliance - Ensures Data Anonymization, Facilitates Secure Testing Speed and Efficiency - Rapid Prototyping, Scalability, Data Quality and Diversity
GenAI Use Case: Code Generation Code Generation [Code generation based on specs] ● Coding Assistance ● Spec-to-code ● Adaptable to industry- specific requirements ● Personalized code generation based on project need ● Offers real-time code suggestions ● Acts as a virtual pair programmer Generate an example of Python code to list prime numbers that are less than 1000 Business Value: Cost-Effectiveness Reduces development time, Lowers costs by automating repetitive tasks Accelerating Development- Rapid prototyping Automation and Scalability- Facilitates Automated Testing, Enables Batch Code Generation # Python code to list prime numbers less than 1000 def is_prime(n): """Check if a number is prime.""" if n <= 1: return False for i in range(2, int(n ** 0.5) + 1): if n % i == 0: return False return True # List prime numbers less than 1000 prime_numbers = [n for n in range(2, 1000) if is_prime(n)] prime_numbers[:10], len(prime_numbers) # Display the first 10 prime numbers and the total count
GenAI Use Case: Code Quality Check Code Quality Checks [Making sure that code does not have security holes or potential bugs] ● Automated code reviews ● Code coverage ● Bug detection ● Unit test cases ● Security vulnerability checks Find bugs in this function … Business Value: Cost-Effectiveness - Reduces development time, Lowers costs by automating repetitive tasks Accelerating Development- Rapid debugging Automation and Scalability- Facilitates Automated Testing, Enables Code Testing as a part of deployment pipeline 1. Bug in is_prime Condition: The condition if n < 1: should be if n <= 1:. The function should return False for both 0 and 1, as they are not prime numbers. 2. Bug in for Loop Range: The loop for i in range(10, int(n ** 0.5) + 1): starts from 10, which will not properly check for the primality of the numbers. The loop should start from 2. 3. Bug in List Comprehension Range: The list comprehension prime_numbers = [n for n in range(0, 10) if is_prime(n)] starts from 0. It should start from 2, as 0 and 1 are not prime numbers. 4. Bug in Output: The output prime_numbers[:1], len(prime_numbers) only shows the first element of the list and the length of the list. If you want to show the first 10 elements, it should be prime_numbers[:10]. 5. Inconsistent Comment: The comment says "List prime numbers less than 10000," but the code is actually listing prime numbers less than 10. This can lead to confusion. 6. Performance Issue (Not exactly a bug): The function could be more efficient if it were to stop checking for divisibility once it finds a divisor.
YOU WON’T BE REPLACED BY AI, YOU’LL BE REPLACED BY SOMEONE WHO KNOWS HOW TO USE AI ● - Harvard Business Review
Next Steps https://www.linkedin.com/learning/generative-ai-for-business-leaders/generative-ai-for- business-leaders?u=2217474 https://www.linkedin.com/learning/generative-ai-for-business-leaders/generative- ai-use-cases?u=2217474 https://www.linkedin.com/learning/generative-ai-working-with-large-language- models/learning-about-large-language-models?u=2217474
QUESTIONS?