Appendix

Appendix: interview summary

The consolidated interview summary behind chapter three.

A summary of our interviews with youth climate negotiators. We asked about the problems they face and how language models could help.

Who we spoke to. Six negotiators, from Liberia, Paraguay, Peru, Nigeria, Lebanon, and Indonesia. Their experience varied, and some were new to the process.

Challenges and pain points

In the role. English is not their first language. UNFCCC jargon is hard (Liberia, Paraguay). Technical knowledge is spread unevenly and poorly shared. They know little about past negotiations (Lebanon). Talking with senior negotiators about complex topics is hard (Peru, Nigeria). Positions in the room change quickly.

In specific situations. Their first COP was daunting because they had not been prepared (Paraguay, Peru). Speaking in public, with so much at stake, makes complex topics hard to explain.

In finding information. The UNFCCC website is the main source, and it is hard to use. There is not enough time to take in key information (Paraguay). News travels through internal messages and personal contacts.

In communicating. Explaining complex ideas quickly and accurately in English is hard. People speak English at very different levels, which gets in the way (Indonesia).

Tools

In use. Grammarly for writing (Paraguay). Google Drive for working on documents together (Paraguay). No translation tools, because negotiation language is hard to translate (Paraguay).

Wished for. Better language and grammar tools made for UNFCCC texts (Paraguay, Liberia). Tools that scan documents quickly and pull out the key points (Lebanon, Nigeria). Learning materials suited to each negotiator (Liberia).

Information and decisions

Staying up to date. Through colleagues and shared material, and through newsletters and websites (Liberia, Lebanon).

Handling complex information. Working as a team and drawing on other people's expertise (Peru). Researching unfamiliar topics in depth.

Working together. Negotiators build supportive communities on purpose (Peru, Nigeria). Efforts to improve language skills are under way (Indonesia).

Ideas and solutions

What they want improved. Simpler official documents, so more people can take part (Lebanon, Nigeria).

Repetitive tasks worth automating. Finding details from past negotiations (Lebanon).

Looking ahead. They see AI and digital tools as possible helpers in negotiations.

Opportunities for language models

Benefits. Translation and summaries, which help with language barriers and long documents.

Automation. Pulling the main points out of long documents.

Concerns. Relying on the technology could weaken people's own research skills (Liberia, Paraguay).