From 8 days to 2 hours: how a scientist who did not trust AI ended up “unable to work without it”
Shahrzad is a PhD candidate at York University, currently working through her comprehensive exams. The format gives her two topics she knows nothing about and a hard deadline to build a defensible literature synthesis on each one.
She came to EMET as a skeptic. She had already tried three other AI tools to find papers, and all three had failed the same test. They pulled sources from subreddits and magazines. When she went back and checked the papers they gave her, the information did not hold up. So when she was given access to EMET, she did not open it for about a month.
"I was skeptical getting into EMET. But once I actually started using it, I haven't found any papers where I've gone and cross checked and found it to be misinformation. It's been accurate pretty much 100% of the time."
Eight days of paper hunting, then two hours
For the first comps topic, she did the literature search herself. Eight days of reading to arrive at a shortlist of 20 papers she wanted to build on. For the second topic, she used EMET. It took two hours.
That is the number she leads with, but it is not the part she was most excited about. What she describes is what the eight days were costing her.
"I have more time to be creative with science, because now I'm not doing the admin work of looking for papers. Finding papers is honestly the hardest and most time consuming part of doing research. Once I have all my papers, I know the research gap, I can come up with the study or experiments and perform them."
She estimates EMET has accelerated her program by at least four months a year.
Trust, built one citation at a time
The thing that moved her was not speed. It was being able to check.
"With EMET, I could literally trace back to the citation. It's right there. And then it's accurate every time."
She still cross checks. She describes herself as new to AI and skeptical, and she treats every claim as something to verify. What she found is that verification kept confirming the tool, and occasionally corrected her.
"Sometimes I'll go and interpret something from a paper, but it interpreted it differently from me. And then I realized my own interpretation may have been wrong. So it's almost like having a collaborator with you at all times."
The confidence to write "to our knowledge, we are the first"
Every scientist knows the discomfort of that sentence. You write that no studies exist, and you hope you searched hard enough.
"Making those vague statements, I usually feel really uncomfortable doing, because I'm always thinking maybe there is a paper that talks about it that I just couldn't find within my own search. With EMET I feel comfortable saying that, because - within my experience - it looks through every single paper out there, even the ones I don't have access to through my university library. So I feel more confident making those big statements."
EMET reaches closed access literature her institution does not subscribe to. When she finds a citation worth keeping, she saves it directly to a project, which replaced the manual OneNote system she had been maintaining by hand.
The scientist maintains full autonomy and creativity
Shahrzad has not handed anything over. She still reads the papers. She still decides which ones align with her question and where the gap is. She has not yet tried EMET on protocol optimization, because previous tools gave her suggestions that were either obvious or wrong, and she wants to test it on her own terms first.
That is the point. The judgement stays with her. What changed is how much of her week is spent on judgement instead of searching. This is the same scientist who left the tool unopened for a month because she did not believe it would work.
"Now I'm addicted. I don't know how I'd live without it, to be honest. I just don't think you could be as optimal as a researcher without this tool."
Her one regret is the month she left it unopened.
