lecture1

1.md

@@ -2,6 +2,14 @@
 ## 1. Large language models (LLMs), Artificial Intelligence (AI) and Natural language processing (NLP), history of NLP
 
 
+>  Explanations, formulas, visualisations: 
+> - Britannica entry for [artificial intelligence](https://www.britannica.com/technology/artificial-intelligence) 
+> - Eisenstein 1.1
+> - Jurafsky-Martin [2](https://web.stanford.edu/~jurafsky/slp3/2.pdf) 
+
+
+
+
   
 
 
@@ -12,17 +20,40 @@
 <img src="figures/NLP_inter.png" alt="inter" width="250"/>
 
 
+- NLP is an important part of artificial intelligence because language is the most prominent capacity of human intelligence. 
+- Another reason why language is central to AI is the famous Turing test.
+- Linguistics was central to language processing in the past, before the introduction of machine learning. Currently we see a big gap between linguistics and NLP, but the potential for bridging this gap is still there.
+- Language processing was one of the first application of machine learning and it remains a prominent challenge.
 
 
-- Computational Linguistics is more about text **parsing**, NLP more about **end user** tasks. 
-- Machine learning is a set of methods and techniques used to parse text and perform end-user tasks.
-- End-user tasks are an important part of artificial intelligence because language is the most prominent capacity of human intelligence. 
 
-(More explanations in Eisenstein)  
+&nbsp; 
 
 
+### Human vs. machine language competence 
+
 &nbsp; 
 
+<img src="figures/Meaning.png" alt="hist" width="550"/>
+
+[Source](https://www.researchgate.net/publication/226728046\_The\_Role\_of_Logic_and_Ontology_in_Language_and_Reasoning)
+
+
+&nbsp; 
+
+<img src="figures/ELIZA.png" alt="hist" width="550"/>
+
+
+[Source](https://web.stanford.edu/~jurafsky/slp3/2.pdf)
+
+
+- Humans use language to name (maybe even to form) concepts by abstracting from real objects and events
+- Machines were able to imitate human language without any access to concepts and reality
+- Modern machines do develop some concepts, but the relationship with the reality is still missing
+
+
+&nbsp;
+
 
 ### Three big eras in the history of NLP