Write a 2,500–4,000 word research paper on current findings in AI, with a defensible angle, primary-source citations, and a section on what it means for builders.
XenonDev builds AI-driven software, and we keep an internal library of research notes our team uses to stay current. We're commissioning a short research paper on the state of AI today — the kind of document a new engineer on our team could read in one sitting and walk away knowing what's actually moving the field.
Deliverable: one research paper, 2,500–4,000 words, covering current findings in AI. We want a clear point of view, not a Wikipedia summary. Pick a defensible angle (for example: where LLM reasoning is breaking down, what's real vs. hyped in agent frameworks, or where open models have closed the gap with closed ones) and argue it with citations.
Structure we expect:
- An intro that says what you're arguing and why it matters.
- A literature section grounded in primary sources from the last 18 months — papers, lab releases, credible benchmarks. No blog-post-of-a-blog-post chains.
- A findings section with your synthesis, not just summaries.
- A short "what this means for builders" section, written for an engineering audience.
- A references list in a consistent format.
What you'll get from us: a starter reading list of about 15 papers, access to a Slack channel with our engineering team for questions, and one 30-minute call with Shaheer mid-way through to pressure-test your angle. Final paper is due by the deadline as a PDF.