Generative AI
Uploaded by
Sai RohithGenerative AI
Uploaded by
Sai RohithGENERATIVE
AI
ECSE BATCH-1
TEAM3
ANUMALASETTY V S
UNDER THE
ROHITH
24071A0509
SUPERVISION OF
N.PRAVE
EN
B.PRATUS BANDA
HA SRIRAM
24071A
TEAM :
OUR
0510
BANDI
RIYA
24071A
0511
BANOTH
DINESH RAJ
24071A
0512
CONTE
0 NTS
WHAT IS
1
0 AI?
TYPES
2
0 OF WHAT
AI IS
3
0 GENERATIVE AI? OF
APPLICATIONS
4
0 GENERATIVE
ETHICAL AI
5
0 CONSIDERATION
GENERATIVE AI IN NATURAL LANGUAGE
6
0 PROCCESING
FUTURE OF
7 GENERATIVE AI
WHAT IS
AI?
• ARTIFICIAL INTELLIGENCE(AI) IS A MULTIDISCIPLINARY FIELD OF
COMPUTER SCIENCE, ALGORITHMS, AND SOFTWARE CAPABLE OF
• SIMULATING HUMAN INTELLIGENCE.
AI CAN SEE, UNDERSTAND, AND TRANSLATE SPOKEN AND
WRITTEN
• AI LANGUAGE
CAN ANALYZE LARGE AMOUNTS OF DATA FROM VARIOUS
SOURCES, SUCH AS SMART SENSORS, MONITORING TOOLS, AND
HUMAN-GENERATED CONTENT
• AI CAN MAKE RECOMMENDATIONS AND SOLVE
COMPLEX PROBLEMS
• AI CAN LEARN AND REASON TO
SOLVE PROBLEMS
TYPES
OF
AI IS AI
ALSO CLASSIFIED INTO
VARIOUS TYPES :
• Tradition
al AI
• Generativ
e AI
TRADITIONA
• TRADITIONAL
L AI : AI OPERATES WITHIN A
• SET
ANDOF RULES
IT IS FOCUSED ON OPTIMIZING TASKS RATHER THAN
EVOLVING CREATIVELY
WHAT IS
INGENERATIVE AI ?
THE CONTEXT OF AI, GENERATIVE ALGORITHMS AND MODELS
HAVE ONES THAT CAN PRODUCE NEW DATA INSTANCES BASED ON
TRAINING EXAMPLES.
• WHILE TRADITIONAL AI OFTEN TARGETS
CLASSIFICATION OR PREDICTION TASKS, GENERATIVE
• AI IS LOOKING TO
GENERATIVE AI GENERATE NEW OUTPUTS.
IS A FORM OF ARTIFICIAL
INTELLIGENCE ABLE TO GENERATE NEW CONTENT
• LIKE TEXT, IMAGES, MUSIC,
CODE AND MORE)
• IT IS TRAINED ON HUGE DATASETS THROUGH ALGORITHMS THAT
ENABLE THEM TO DISCOVER PATTERNS AND STRUCTURES WITHIN
THE DATA
• WHEN IT WAS TRAINED, IT CAN GENERATE NEW CONTENT THAT
RESEMBLES THE TRAINING DATA BUT IS DIFFERENT
APPLICATIONS OF
GENERATIVE AI
ETHICAL
CONSIDERATION
01. DATA
PRIVACY
DATA - LONG AND CONTAINS MORE INFORMATIVE DATA SO
CAN BE VERY
THESE GENERATIVE AI ALGORITHMS ARE TRAINED ON TOP OF
THAT.RANGE OF PRIVACY ATTACKING MODELS MAY ABLE TO LEARN AS A
RESULT FROM THIS THING. WITHOUT CARE, OR IF USED IMPROPERLY,
THIS CAN BE A HUGE PROBLEM.
02.BIAS AND
FAIRNESS
THE IDEAL - WHICH ARE FREE OF ANY BIAS OR
AI SYSTEMS
DISCRIMINATION CAUSING NO HARM TO ANYONE. BUT IT DOES
NOT GUARANTEE THAT THE AI ALGORITHM IS BIASED TOWARDS
THE PREFERENCES OF ITS CREATORS OR DATA.
03.
TRANSPARANC
IN ORDER TO AVOID UNMET EXPECTATIONS IT IS CRUCIAL TO
SET THE RIGHT EXPECTATION WITH YOUR CLIENTS AND
Y -
STAKEHOLDERS ON WHAT THEY CAN EXPECT FROM AI-
POWERED TOOLS. THIS WILL INVOLVE EXPLAINING WHAT THE
AI DOES FOR CARE, TALKING ABOUT THE PROS AND CONS, AND
RESOLVING04.
DOUBTS.
AI SYSTEMS SHOULD BLAME EITHER BEFALL ON DEVELOPERS
ACCOUNTABILIT
OR END-USERS OF THE AI ENDPOINTS. TO DO SO, IT IS
NECESSARYY NOT
- TO USE AI SYSTEMS IRRESPONSIBLY.
05.. COPYRIGHT AND
UNCERTAINTY ABOUT THE AUTHORSHIP AND COPYRIGHT OF AI
AUTHORSHIP -
GENERATED CONTENT. IT DEFINES THE OWNERSHIP OF CREATIVE
WORKS AND THE CONDITIONS OF THEIR USE.
