Please enable JavaScript.
Coggle requires JavaScript to display documents.
Ds Task - Coggle Diagram
Ds Task
Alignment
We have to understand the values and morals and views of the artificial intelligence should match with our own morals and vales and morals.
-
Compute
The amount of computing process needed to create the artificial intelligence or to train the artificial intelligence is significantly more.
Diffusion
Creation of images. and they do that by destroying their training data and as well as also being able to reverse the process.
-
Foundation Models
Advanced AIs capable of a range of skills, raising questions about factual accuracy and biases.
-
-
-
-
-
-
-
Neural Network
Are interconnected nodes that mimic the human brain, allowing AI to learn and make predictions based on patterns and connections.
-
Prompt engineering
involves writing effective prompts to elicit desired responses from AI systems, potentially becoming a valuable job skill.
Quantum machine learning
Explores the combination of quantum processes and machine learning techniques to enhance AI's computation and generalization abilities.
The "race to the bottom"
Concerns the potential negative impacts of rapidly advancing AI without proper safeguards, regulation, and ethical considerations.
-
Superintelligence
Refers to machines surpassing human mental capabilities, raising questions about alignment, emergent behavior, and potential risks.
Training data
Is crucial for AI to learn and make predictions, and its quality, biases, and size influence AI performance and accuracy.
Unsupervised learning
Allows AI to learn from unlabelled data and form its own concepts, potentially leading to more knowledgeable and accurate AI.
Voice cloning
Uses AI to create digital voices that sound like real individuals, raising ethical questions about consent and authenticity.
Weak AI
Refers to AI systems that perform well in specific tasks but lack the general intelligence of humans.
-
X-risk
Refers to the existential risk posed by AI, where its development and potential impacts on humanity should be carefully regulated and controlled.
-
Zero-shot learning
Allows AI to respond to concepts or objects it has never encountered before by applying its knowledge and inference abilities.