I was thinking the same. A lot of people use AI to refine their writing and make it more concise.
I do think the dead giveaways (em dashes, it's not X it's Y, etc.) are annoying to come across repeatedly. A person not bothering to remove these tells feels 'low effort' to me.
Great outside-the-box thinking, will be giving this a try, especially with LLMs having computer control functionality. Combine this with some open source video editors and you've basically got your own bootstrapped version of Screen Studio.
Very cool use of AI. Shows the power of Autoresearch when you're actually able to build a strict set of tests that clearly delineate successful outcomes.
One of the things they neglected to discuss was how much longer it takes given the synthesis step. I guess for a deep research benchmark, it doesn't matter much, but will be interesting to see how it applies to coding tasks.
Based on your explanation, it doesn't sound feasible for me, a complete non-engineer, to switch to fully offline? I do a lot of back and forth discussion with LLMs as someone who reads and writes 0 code.
Do people even use the visual/graph views at all or is it just optics? I'm thinking about it from the perspective of using LLMs to search your own database, and I imagine it reaches a scale where it's not feasible or effective for a human to be manually searching.
I do think the dead giveaways (em dashes, it's not X it's Y, etc.) are annoying to come across repeatedly. A person not bothering to remove these tells feels 'low effort' to me.