Hacker News Reader

This is fun to do, but you’ll rapidly find that the academic sources don’t even say what they claim to.

The foundations of modern science seem to be quite shaky...

Nothing is stopping anyone with access to the SciHub corpus from doing this.

It would be useful. You'd also hit a new paywall for each journal, and the tool would end up costing thousands of dollars for anyone to use. You could of course I terrace with scihub, imo one of the most important tools in the last 20 years, but then you can't build a business on it

Although you could infer bad studies by citations, external criticisms etc, ultimately wouldn’t you have to look at the content of the studies themselves? The ubiquitous paywalls in online academic repos seems like a big obstacle to this.