Try Devin on research tasks (mostly failed)
From the linked report, answer.AI gave Devin, which works through Slack in its own Docker environment with a browser, editor and shell, two tasks. One was grabbing an The builder's 25 most recent Google Scholar papers with Playwright, skipping paywalls. On the scraping task Devin got stuck parsing HTML and went to sleep.
- Agent or tool
- Devin
- Topic
- Research & Data
- Reported on
- Blog
- Added
- 2026-10-15
Why it matters
Research is mostly gathering, cleaning and comparing. Handing that groundwork to an agent leaves the person with the part that needs judgment: deciding what the findings mean. It is one of 201 sightings filed under Research & Data on Agent Sightings (number 182 by interest), and one of 20 involving Devin.
How you could try it
Here is a reusable prompt, written by us in our own words, that captures the task. Paste it into Devin or any similar agent and adjust the details to your situation.
Research brief: try Devin on research tasks (mostly failed). Prefer primary sources, include links and confidence notes, and do not contact anyone.
Tip: Start small: run it once on a low-stakes example, read what the agent did, and only then let it run on real work.
Source
Read the original first-hand account on Blog: https://www.answer.ai/posts/2025-01-08-devin.html. Our summary is a paraphrase; details are in the source.
Built an AI agent? Feature your agent on Agent Sightings, free submission or a featured spot.
Browse more like this
Related sightings
- 50 AI agents in Minecraft built a government, a ration ledger and a no-blame inquiry about a chicken
- Pensive offers personalized AI teaching assistants for higher ed with Claude
- From weeks to a day: Everstar on Microsoft Foundry with Claude
- Biomni accelerates biomedical discoveries by 100x with Claude
- Async Devin chores between morning and lunch
- Triage client messages into tickets and replies