Activities worth dedicating time in the age of artifical intelligence

Things worth spending time with:

• requirement engineering: identifying scope and system boundaries
• team work: improving development methodologies
• strategic planning: making decisions, selecting the right tools
• tooling: knowing how to utilise tools, eg. how to apply AI
• ethics
• functional safety
• evaluating text meaningfulness, identifying AI generated text

Things AI can do or will do soon:

• writing tutorials and getting-started guides
• performing basic data exploration
• detecting anomalies
• creating filler text
• dealing with legal text: summarizing, writing

Where to begin

Algorithms: Lin Reg -> Log Reg -> NN -> CNN + RNN -> GANs + Transformers -> ViT -> Multimodal AI + LLMs + Diffusion + Auto Encoders SVM, PCA, kNN, k-means clustering, etc.

LightGBM, XGboost, Catboost, etc.

Optimization and optimizers.

Application-wise: Classification, Semantic Segmentation, Pose Estimation, Text Generation, Summarization, NER, Image Generation, Captioning, Sequence Generation (like music/speech), text to speech, speech to text, recommender systems, sentiment amalysis, tabular data, etc. - Source

Resources:

• Chris Albon
• Neural networks and deep learning
• Landmark papers
• Introduction to Statistical Learning

AI alignment

• Faulty reward functions
• Goodhart's law