Research Papers and a Visa Fiasco
_arohan_ · x · 2026-07-12
The author recounts leading a research paper for the first time, aiming to accelerate distributed training by enabling inter-node communication—a approach previously successful in Google's Search and Ads pipelines.
This quickly turns into a dramatic visa fiasco: he originally planned to attend ICLR in Vancouver but got rejected. He later realized the rejection letter was addressed to someone else, and the system could actually handle this. Geoff Hinton tried to help but was advised not to intervene. After a month of investigation, they discovered the real issue wasn't a rejection—the visa was actually approved, but the stamp was only valid for 30 days. Furthermore, the process required mailing the passport to one location and the application fee check to another state. Ultimately, he had to catch the earliest flight to Los Angeles, stay in a hotel overnight, and race against time to submit the documents.
More from Research
- SUFLECA shows NOC-based correspondence can improve CAD-to-image alignment — ducha_aiki · 2026-07-21
- OpenAI-style autonomous researchers could become real scientific collaborators — Promptmethus · 2026-07-21
- Soft Clamp cuts tool-call overuse in multi-teacher distillation, from 13.7% to 9.0% — antgroup · 2026-07-21
- ShotPlan adds learnable planning tokens for cinematic multi-shot video generation — Tele-AI · 2026-07-21
- A silicon photonic reservoir chip compensates fiber distortion in real time at 28 Gbps — bravo_abad · 2026-07-21
- A developer maps out six design rules for CLIs that humans and AI agents can both use — yujiezha · 2026-07-21