Literature
Natural language generation
AI & Writing
Also known as Generative artificial intelligence, Large language model, AI writing
When a machine can produce smooth, fluent text on command, it throws open old questions: who counts as the author, what makes writing authentic, and what words are even worth. This collides with law, where rules on ownership and credit lag behind a tool that can outrun them. It is a live debate in philosophy about how our tools quietly reshape the way we think and who we are. And it worries media, since a flood of cheap, convincing text makes it far harder to tell real information from the fake.
Key people
- Margaret MitchellResearcher, natural language processing
- John A. BatemanBritish linguist and semiotician
- Hady ElsaharResearcher in Natural Language Processing and Machine Learning
Timeline
- 1990NLG has existed since ELIZA was developed in the mid 1960s, but the methods were first used commercially in the 1990s.
Read
- The Laws of Human NatureRobert Greene · 2018Book
- Natural Language Processing With PythonEdward Loper · 2009Book
- Discourse Function and Syntactic Form in Natural Language GenerationCassandre Creswell · 2004Book
- Natural Language Generation in Interactive SystemsAmanda Stent · 2014Book
Listen
- Natural Language GenerationKeelin MPodcast
- IFTTD - If This Then DevBruno Soulez | Orso MediaPodcast
Voices to follow
- Trace NicholsYouTubeAmerican visual artist, photographer, and educator
- Emanuel GollobYouTubeAustrian artist and researcher
- Erwan Soumhi@ErwanSoumhi · XFrench conceptual artist, film director and AI visual anthropologist
Debates
- Can NLG truly achieve human-level creativity?One view: Yes, advanced models can generate novel and complex outputs that are often indistinguishable from human work. · Another: No, current NLG relies on pattern recognition and lacks genuine understanding or consciousness, limiting true creativity.Open question
- Should NLG prioritize factual accuracy over engaging language?One view: Factual accuracy is paramount for trustworthy information, especially in critical applications like news or medical texts. · Another: Engaging language is vital for effective communication and user experience, even if it means slight compromises for readability.Open question
- How much human oversight is necessary for NLG systems?One view: Extensive human oversight is crucial to prevent bias, misinformation, and misuse in generated content. · Another: As NLG improves, systems can become more autonomous, reducing the need for constant human intervention and increasing efficiency.Open question
Glossary
- Natural Language Generation (NLG)The use of artificial intelligence to produce human-like text from structured data or other inputs.
- Large Language Model (LLM)A type of AI model trained on vast amounts of text data to understand, generate, and respond to human language.
- Prompt EngineeringThe skill of crafting effective input instructions (prompts) to guide an AI model to produce desired outputs.
- Natural Language Understanding (NLU)The branch of AI focused on enabling computers to comprehend and interpret human language.
- TokenA fundamental unit of text, such as a word, part of a word, or punctuation mark, that an AI model processes.
- HallucinationWhen an NLG model generates false, nonsensical, or unfaithful information with high confidence.
Careers
Roles this can lead toward
Student research
Published policy papers by One Young India delegates — every delegate leaves published under their own name.
- Language in Education Policy: For a Multilingual IndiaAnvesha Kaustubh Kukde
- The Decline of Regional Languages in India – Preserving Linguistic Diversity in a Globalized EraSajinkya Sharan Gupta
Threads 3
Where this connects to other fields — and why it's worth knowing.
- AI & Technology Regulation Law
Copyright law was built on one assumption: a human wrote it. AI that writes poems and stories breaks that assumption in half. Now judges have to decide brand-new questions, like whether a machine's words can even be owned, and whether feeding it thousands of books to learn from counts as stealing.
- Philosophy of Technology Philosophy
Plato griped that writing would wreck people's memories and hand them fake wisdom instead of real understanding. That's word-for-word the panic we now aim at AI. The fear that 'this new tech will ruin our minds' is at least 2,400 years old, which should make us pause before repeating it.
- Misinformation & Disinformation Media
When a machine can spit out endless smooth, human-sounding text for free, lying becomes basically costless: you can flood the internet with fakes in seconds. So the thing that suddenly gets rare and precious isn't information, it's trust. Once anyone can fake anything, knowing who to believe becomes the whole game.