SlideShare a Scribd company logo
OpenAI
(ChatGPT)
Multimedia
and Modern
Technology
6
A.L. Ridwan Boya Marqas
Computer Science Dept.
Nawroz University
2022 - 2023
OpenAI
– OpenAI is an American artificial intelligence research
laboratory consisting of the non-profit OpenAI
Incorporated and its for-profit subsidiary corporation
OpenAI Limited Partnership. OpenAI conducts AI
research with the declared intention of promoting
and developing a friendly AI.
ChatGPT
– As an AI language model, ChatGPT
is a program consisting of a large
neural network that has been
trained on vast amounts of textual
data. Specifically, ChatGPT is a
variant of the GPT (Generative Pre-
trained Transformer) family of
models developed by OpenAI.
ChatGPT
–ChatGPT consists of many layers of artificial neurons, which are
interconnected in a way that allows the model to process and
analyze text data. The model is pre-trained on a large corpus of
text data, such as books, articles, and web pages, to learn
patterns and relationships between words, phrases, and
sentences. This pre-training allows ChatGPT to generate coherent
and contextually appropriate responses to user input, even if it has
ChatGPT
–ChatGPT also incorporates techniques such as
attention mechanisms, which allow the model to
focus on different parts of the input text when
generating responses, and fine-tuning, which allows
the model to be further trained on specific tasks or
domains.
ChatGPT Algorithm
–ChatGPT is a deep learning model that uses a specific type of
algorithm called a transformer. Specifically, ChatGPT uses a
variant of the transformer algorithm that has been modified and
optimized for natural language processing (NLP) tasks. This
variant of the transformer is called the GPT (Generative Pre-
trained Transformer) algorithm.
ChatGPT Algorithm work
–The GPT algorithm is a type of neural network that uses attention
mechanisms to process and generate natural language text. It works by
breaking down a sequence of input text into smaller sub-sequences,
called tokens, which are then processed by multiple layers of artificial
neurons. Each layer of neurons applies a non-linear transformation to
the input tokens, and attention mechanisms are used to allow the model
to selectively focus on different parts of the input when generating
responses.
ChatGPT Algorithm Work
–The training of ChatGPT and other GPT-based models involves a
process called unsupervised learning, where the model is trained
on large amounts of unstructured text data. This training allows the
model to learn patterns and relationships between words, phrases,
and sentences, and to generate contextually appropriate
responses to user input.
ChatGPT detectors
– GPT-Zero
– OpenAI GPT2 output detector
– Hello-SimpleAI
– Contentatscale AI content detector
– Writers AI content detector
GPTZero
– GPTZero is designed by Edward Tian, a Princeton senior, to
detect writing generated by ChatGPT. The creator has seen that
the technology was used by students to cheat on assignments,
so he came up with the safeguard.
OpenAI GPT2 output detector
– OpenAI GPT2 Output Detector is developed by OpenAI and hosted on
HuggingFace for users to freely check the content validity. It can
detect text generated by ChatGPT, GPT3, and GPT2. The application
is using the GPT-2 output detector model based on the Transformers.
Hello-Simple AI ChatGPT
detector
– Hello-SimpleAI ChatGPT Detector consists of three versions to detect text
generated with ChatGPT.
– QA version: requires both questions and answers to check if it was generated
by ChatGPT. On the backend, it is using a PLM-based classifier.
– Sinlge-text version: required long text to detect text generated by ChatGPT
by using PLM-based classifiers.
– Linguistic version: detects ChatGPT based on linguistic features
Contentatscale AI content
Detector
– Contentatscale AI Content Detector is the tool that I frequently use for
quality assessment. It is fast, simple, and accurate. The AI Content
Detector allows users to receive the Human Content score by pasting
the text.
Writers AI Content Detector
– Writers AI Content Detector is similar to Contentatscale. It requires
either the URL of the page or text to calculate the “Human-Generated
Content” score. It is fast and fairly accurate
OpenAI ChatGPT Passed Advanced
Exams
– US medical licensing exam
– Essays
– Microbiology quiz
– Law School Exams
– Stanford Medical School clinical reasoning final
ChatGPT limitations
– Lack of common sense
– Lack of emotional intelligence
– Limitations in understanding context
– Trouble generating long-form, structured content
– Limitations in handling multiple tasks at the same time
– Potentially biased responses
– Limited knowledge
– Accuracy problems or grammatical issues
– Need for fine-tuning
– Computational costs and power
ChatGPT Capabilities
1. Generate Email
Templates
2. Use as Search
Engine
3. Explain in
Different Styles
4. Brainstorm Ideas
5. Make
Conversations
About Your Favorite
Topic
6. Complete Your
Essays and
Assignments
7. Generate Jokes
8. Play Trivia
Games
9. Treat It Like a
Mentor
10. Generate Bio
and Posts for Social
Media
11. Translate Text
12. Extract Data
From Text
13. Summarize
Articles
14. Generate
YouTube Scripts
15. Write Blog
Posts
16. Play Guessing
Games
17. Write Code 18. Debug Code
19. Ask for Movie
and Book
Recommendations
20. Generate Movie
Scripts and Stories
21. Write Song
Lyrics and Poems
22. Get Random
Facts
24. Complete
Crossword Puzzles
25. Create an App
OpenAI Chatgpt.pptx
OpenAI Chatgpt.pptx
OpenAI Chatgpt.pptx
OpenAI Chatgpt.pptx
OpenAI Chatgpt.pptx
OpenAI Chatgpt.pptx
OpenAI Chatgpt.pptx
OpenAI Chatgpt.pptx
OpenAI Chatgpt.pptx
OpenAI Chatgpt.pptx
OpenAI Chatgpt.pptx
OpenAI Chatgpt.pptx
OpenAI Chatgpt.pptx
OpenAI Chatgpt.pptx
OpenAI Chatgpt.pptx

More Related Content

What's hot (20)

PPTX
Webinar on ChatGPT.pptx
Abhilash Majumder
 
PDF
ChatGPT-the-revolution-is-coming.pdf
Liang Yan
 
PPTX
What Is GPT-3 And Why Is It Revolutionizing Artificial Intelligence?
Bernard Marr
 
PDF
AI and ML Series - Introduction to Generative AI and LLMs - Session 1
DianaGray10
 
PDF
Let's talk about GPT: A crash course in Generative AI for researchers
Steven Van Vaerenbergh
 
PPTX
How ChatGPT and AI-assisted coding changes software engineering profoundly
Pekka Abrahamsson / Tampere University
 
PDF
The Rise of the LLMs - How I Learned to Stop Worrying & Love the GPT!
taozen
 
PPTX
ChatGPT 101.pptx
MohamadAimanArifMoha
 
PDF
Using the power of Generative AI at scale
Maxim Salnikov
 
PPTX
Jawad's presentation on GPT.pptx
JawadNadeem3
 
PPTX
ChatGPT ppt.pptx
YuvrajS9
 
PPTX
Episode 2: The LLM / GPT / AI Prompt / Data Engineer Roadmap
Anant Corporation
 
PPTX
200109-Open AI Chat GPT-4-3.pptx
andre241421
 
PPTX
5 BENIFITES OF CHAT GPT.pptx
shailesh sangle
 
PPTX
chatGPT.txt
Kamleshlodhi1
 
PPTX
How to Teach and Learn with ChatGPT - BETT 2023
Dominik Lukes
 
PDF
An Introduction to Generative AI - May 18, 2023
CoriFaklaris1
 
PDF
ChatGPT Evaluation for NLP
XiachongFeng
 
PDF
ChatGPT 101 - Vancouver ChatGPT Experts
Ali Tavanayan
 
PPTX
Introduction to ChatGPT
Damian T. Gordon
 
Webinar on ChatGPT.pptx
Abhilash Majumder
 
ChatGPT-the-revolution-is-coming.pdf
Liang Yan
 
What Is GPT-3 And Why Is It Revolutionizing Artificial Intelligence?
Bernard Marr
 
AI and ML Series - Introduction to Generative AI and LLMs - Session 1
DianaGray10
 
Let's talk about GPT: A crash course in Generative AI for researchers
Steven Van Vaerenbergh
 
How ChatGPT and AI-assisted coding changes software engineering profoundly
Pekka Abrahamsson / Tampere University
 
The Rise of the LLMs - How I Learned to Stop Worrying & Love the GPT!
taozen
 
ChatGPT 101.pptx
MohamadAimanArifMoha
 
Using the power of Generative AI at scale
Maxim Salnikov
 
Jawad's presentation on GPT.pptx
JawadNadeem3
 
ChatGPT ppt.pptx
YuvrajS9
 
Episode 2: The LLM / GPT / AI Prompt / Data Engineer Roadmap
Anant Corporation
 
200109-Open AI Chat GPT-4-3.pptx
andre241421
 
5 BENIFITES OF CHAT GPT.pptx
shailesh sangle
 
chatGPT.txt
Kamleshlodhi1
 
How to Teach and Learn with ChatGPT - BETT 2023
Dominik Lukes
 
An Introduction to Generative AI - May 18, 2023
CoriFaklaris1
 
ChatGPT Evaluation for NLP
XiachongFeng
 
ChatGPT 101 - Vancouver ChatGPT Experts
Ali Tavanayan
 
Introduction to ChatGPT
Damian T. Gordon
 

Recently uploaded (20)

PDF
RAT Builders - How to Catch Them All [DeepSec 2024]
malmoeb
 
PDF
The Future of Artificial Intelligence (AI)
Mukul
 
PPTX
Agile Chennai 18-19 July 2025 | Emerging patterns in Agentic AI by Bharani Su...
AgileNetwork
 
PPTX
Using Google Data Studio (Looker Studio) to Create Effective and Easy Data Re...
Orage Technologies
 
PPTX
Simple and concise overview about Quantum computing..pptx
mughal641
 
PPTX
Agile Chennai 18-19 July 2025 Ideathon | AI Powered Microfinance Literacy Gui...
AgileNetwork
 
PPTX
Applied-Statistics-Mastering-Data-Driven-Decisions.pptx
parmaryashparmaryash
 
PDF
Brief History of Internet - Early Days of Internet
sutharharshit158
 
PPTX
Machine Learning Benefits Across Industries
SynapseIndia
 
PDF
GDG Cloud Munich - Intro - Luiz Carneiro - #BuildWithAI - July - Abdel.pdf
Luiz Carneiro
 
PDF
introduction to computer hardware and sofeware
chauhanshraddha2007
 
PDF
OFFOFFBOX™ – A New Era for African Film | Startup Presentation
ambaicciwalkerbrian
 
PDF
Lecture A - AI Workflows for Banking.pdf
Dr. LAM Yat-fai (林日辉)
 
PDF
How ETL Control Logic Keeps Your Pipelines Safe and Reliable.pdf
Stryv Solutions Pvt. Ltd.
 
PDF
Build with AI and GDG Cloud Bydgoszcz- ADK .pdf
jaroslawgajewski1
 
PDF
MASTERDECK GRAPHSUMMIT SYDNEY (Public).pdf
Neo4j
 
PPTX
AVL ( audio, visuals or led ), technology.
Rajeshwri Panchal
 
PDF
TrustArc Webinar - Navigating Data Privacy in LATAM: Laws, Trends, and Compli...
TrustArc
 
PPTX
Farrell_Programming Logic and Design slides_10e_ch02_PowerPoint.pptx
bashnahara11
 
PPTX
PCU Keynote at IEEE World Congress on Services 250710.pptx
Ramesh Jain
 
RAT Builders - How to Catch Them All [DeepSec 2024]
malmoeb
 
The Future of Artificial Intelligence (AI)
Mukul
 
Agile Chennai 18-19 July 2025 | Emerging patterns in Agentic AI by Bharani Su...
AgileNetwork
 
Using Google Data Studio (Looker Studio) to Create Effective and Easy Data Re...
Orage Technologies
 
Simple and concise overview about Quantum computing..pptx
mughal641
 
Agile Chennai 18-19 July 2025 Ideathon | AI Powered Microfinance Literacy Gui...
AgileNetwork
 
Applied-Statistics-Mastering-Data-Driven-Decisions.pptx
parmaryashparmaryash
 
Brief History of Internet - Early Days of Internet
sutharharshit158
 
Machine Learning Benefits Across Industries
SynapseIndia
 
GDG Cloud Munich - Intro - Luiz Carneiro - #BuildWithAI - July - Abdel.pdf
Luiz Carneiro
 
introduction to computer hardware and sofeware
chauhanshraddha2007
 
OFFOFFBOX™ – A New Era for African Film | Startup Presentation
ambaicciwalkerbrian
 
Lecture A - AI Workflows for Banking.pdf
Dr. LAM Yat-fai (林日辉)
 
How ETL Control Logic Keeps Your Pipelines Safe and Reliable.pdf
Stryv Solutions Pvt. Ltd.
 
Build with AI and GDG Cloud Bydgoszcz- ADK .pdf
jaroslawgajewski1
 
MASTERDECK GRAPHSUMMIT SYDNEY (Public).pdf
Neo4j
 
AVL ( audio, visuals or led ), technology.
Rajeshwri Panchal
 
TrustArc Webinar - Navigating Data Privacy in LATAM: Laws, Trends, and Compli...
TrustArc
 
Farrell_Programming Logic and Design slides_10e_ch02_PowerPoint.pptx
bashnahara11
 
PCU Keynote at IEEE World Congress on Services 250710.pptx
Ramesh Jain
 
Ad

OpenAI Chatgpt.pptx

  • 1. OpenAI (ChatGPT) Multimedia and Modern Technology 6 A.L. Ridwan Boya Marqas Computer Science Dept. Nawroz University 2022 - 2023
  • 2. OpenAI – OpenAI is an American artificial intelligence research laboratory consisting of the non-profit OpenAI Incorporated and its for-profit subsidiary corporation OpenAI Limited Partnership. OpenAI conducts AI research with the declared intention of promoting and developing a friendly AI.
  • 3. ChatGPT – As an AI language model, ChatGPT is a program consisting of a large neural network that has been trained on vast amounts of textual data. Specifically, ChatGPT is a variant of the GPT (Generative Pre- trained Transformer) family of models developed by OpenAI.
  • 4. ChatGPT –ChatGPT consists of many layers of artificial neurons, which are interconnected in a way that allows the model to process and analyze text data. The model is pre-trained on a large corpus of text data, such as books, articles, and web pages, to learn patterns and relationships between words, phrases, and sentences. This pre-training allows ChatGPT to generate coherent and contextually appropriate responses to user input, even if it has
  • 5. ChatGPT –ChatGPT also incorporates techniques such as attention mechanisms, which allow the model to focus on different parts of the input text when generating responses, and fine-tuning, which allows the model to be further trained on specific tasks or domains.
  • 6. ChatGPT Algorithm –ChatGPT is a deep learning model that uses a specific type of algorithm called a transformer. Specifically, ChatGPT uses a variant of the transformer algorithm that has been modified and optimized for natural language processing (NLP) tasks. This variant of the transformer is called the GPT (Generative Pre- trained Transformer) algorithm.
  • 7. ChatGPT Algorithm work –The GPT algorithm is a type of neural network that uses attention mechanisms to process and generate natural language text. It works by breaking down a sequence of input text into smaller sub-sequences, called tokens, which are then processed by multiple layers of artificial neurons. Each layer of neurons applies a non-linear transformation to the input tokens, and attention mechanisms are used to allow the model to selectively focus on different parts of the input when generating responses.
  • 8. ChatGPT Algorithm Work –The training of ChatGPT and other GPT-based models involves a process called unsupervised learning, where the model is trained on large amounts of unstructured text data. This training allows the model to learn patterns and relationships between words, phrases, and sentences, and to generate contextually appropriate responses to user input.
  • 9. ChatGPT detectors – GPT-Zero – OpenAI GPT2 output detector – Hello-SimpleAI – Contentatscale AI content detector – Writers AI content detector
  • 10. GPTZero – GPTZero is designed by Edward Tian, a Princeton senior, to detect writing generated by ChatGPT. The creator has seen that the technology was used by students to cheat on assignments, so he came up with the safeguard.
  • 11. OpenAI GPT2 output detector – OpenAI GPT2 Output Detector is developed by OpenAI and hosted on HuggingFace for users to freely check the content validity. It can detect text generated by ChatGPT, GPT3, and GPT2. The application is using the GPT-2 output detector model based on the Transformers.
  • 12. Hello-Simple AI ChatGPT detector – Hello-SimpleAI ChatGPT Detector consists of three versions to detect text generated with ChatGPT. – QA version: requires both questions and answers to check if it was generated by ChatGPT. On the backend, it is using a PLM-based classifier. – Sinlge-text version: required long text to detect text generated by ChatGPT by using PLM-based classifiers. – Linguistic version: detects ChatGPT based on linguistic features
  • 13. Contentatscale AI content Detector – Contentatscale AI Content Detector is the tool that I frequently use for quality assessment. It is fast, simple, and accurate. The AI Content Detector allows users to receive the Human Content score by pasting the text.
  • 14. Writers AI Content Detector – Writers AI Content Detector is similar to Contentatscale. It requires either the URL of the page or text to calculate the “Human-Generated Content” score. It is fast and fairly accurate
  • 15. OpenAI ChatGPT Passed Advanced Exams – US medical licensing exam – Essays – Microbiology quiz – Law School Exams – Stanford Medical School clinical reasoning final
  • 16. ChatGPT limitations – Lack of common sense – Lack of emotional intelligence – Limitations in understanding context – Trouble generating long-form, structured content – Limitations in handling multiple tasks at the same time – Potentially biased responses – Limited knowledge – Accuracy problems or grammatical issues – Need for fine-tuning – Computational costs and power
  • 17. ChatGPT Capabilities 1. Generate Email Templates 2. Use as Search Engine 3. Explain in Different Styles 4. Brainstorm Ideas 5. Make Conversations About Your Favorite Topic 6. Complete Your Essays and Assignments 7. Generate Jokes 8. Play Trivia Games 9. Treat It Like a Mentor 10. Generate Bio and Posts for Social Media 11. Translate Text 12. Extract Data From Text 13. Summarize Articles 14. Generate YouTube Scripts 15. Write Blog Posts 16. Play Guessing Games 17. Write Code 18. Debug Code 19. Ask for Movie and Book Recommendations 20. Generate Movie Scripts and Stories 21. Write Song Lyrics and Poems 22. Get Random Facts 24. Complete Crossword Puzzles 25. Create an App

Editor's Notes

  • #5: يتكون ChatGPT من عدة طبقات من الخلايا العصبية الاصطناعية ، والمترابطة بطريقة تسمح للنموذج بمعالجة البيانات النصية وتحليلها. تم تدريب النموذج مسبقًا على مجموعة كبيرة من البيانات النصية ، مثل الكتب والمقالات وصفحات الويب ، لتعلم الأنماط والعلاقات بين الكلمات والعبارات والجمل. يسمح هذا التدريب المسبق لـ ChatGPT بإنشاء استجابات متماسكة ومناسبة للسياق لإدخال المستخدم ، حتى لو لم يواجه هذا الإدخال المعين من قبل.
  • #6: يشتمل ChatGPT أيضًا على تقنيات مثل آليات الانتباه ، والتي تسمح للنموذج بالتركيز على أجزاء مختلفة من نص الإدخال عند إنشاء الردود ، والضبط الدقيق ، مما يسمح للنموذج بالتدرب بشكل أكبر على مهام أو مجالات محددة.
  • #7: ChatGPT هو نموذج تعلم عميق يستخدم نوعًا معينًا من الخوارزمية يسمى المحول. على وجه التحديد ، يستخدم ChatGPT متغيرًا من خوارزمية المحولات التي تم تعديلها وتحسينها لمهام معالجة اللغة الطبيعية (NLP). يسمى هذا المتغير من المحولات خوارزمية GPT (المحولات التوليدية المدربة مسبقًا). The Transformer algorithm is a type of deep learning model used for natural language processing (NLP). It was introduced in 2017 by Vaswani et al. in their paper "Attention Is All You Need".
  • #8: تعد خوارزمية GPT نوعًا من الشبكات العصبية التي تستخدم آليات الانتباه لمعالجة وإنشاء نص بلغة طبيعية. إنه يعمل عن طريق تقسيم تسلسل إدخال النص إلى تسلسلات فرعية أصغر ، تسمى الرموز ، والتي تتم معالجتها بعد ذلك بواسطة طبقات متعددة من الخلايا العصبية الاصطناعية. تطبق كل طبقة من الخلايا العصبية تحولًا غير خطي على رموز الإدخال ، وتُستخدم آليات الانتباه للسماح للنموذج بالتركيز بشكل انتقائي على أجزاء مختلفة من الإدخال عند إنشاء الاستجابات.
  • #9: يتضمن تدريب ChatGPT والنماذج الأخرى المستندة إلى GPT عملية تسمى التعلم غير الخاضع للإشراف ، حيث يتم تدريب النموذج على كميات كبيرة من البيانات النصية غير المنظمة. يسمح هذا التدريب للنموذج بتعلم الأنماط والعلاقات بين الكلمات والعبارات والجمل ، وإنشاء استجابات مناسبة للسياق لإدخال المستخدم.
  • #17: عدم وجود الحس السليم نقص الذكاء العاطفي القيود في فهم السياق مشكلة في إنشاء محتوى منظم طويل الشكل القيود في التعامل مع مهام متعددة في نفس الوقت ردود يحتمل أن تكون متحيزة معرفة محدودة مشاكل الدقة أو المسائل النحوية الحاجة إلى صقل التكاليف والقوة الحاسوبية