Been trying to think about what next session themes could be useful for the society. In the session yesterday we talked about if the 'scientific paper' as as form was optimal + what the next scientific paper could look like. What would the future be of scientific communication?
The other idea was a session on how to write more efficiently. Whilst I think the process of writing is the most important way to reason about the world, and one which shouldn't be delegated to AI, there are still lots of little things that make your life easier if you do delegate them. For example, inserting wikipedia images, cleaning up code blocks, writing up scripts to cross post things, and cleaning up reference formatting. One thing indep scientists could do to have an edge is to write much clearer and have broader outreach, and a lot of indeps have this edge because we are forced to do things in much shorter time spans because of our day jobs.
Society session ideas
-
casualphysicsenjoyer
- Site Admin
- Posts: 12
- Joined: Sat Jun 20, 2026 8:35 am
Re: Society session ideas
Tyler Cowen wrote about the future of research papers (link: https://marginalrevolution.com/marginal ... omics.html)
The obvious value of this (to me) would be if experiments were laid out very clearly for people to replicate + publish their own results + play around with different controls + a place for people to add their own results/method etc to the same question(s) addressed in the paper
Some of what he talks about:
"Soon enough you will be able to take any published research paper and tweak it, or improve it, any way you want"
"Why not write a program, or have an AI write it for you, that will take your favorite papers and improve them, and change their evaluations over time, as new results come in?"
"Will highly productive researchers, who publish a lot of papers, become far less valuable? The individual paper no longer seems scarce, or will not be in another year or two."
The obvious value of this (to me) would be if experiments were laid out very clearly for people to replicate + publish their own results + play around with different controls + a place for people to add their own results/method etc to the same question(s) addressed in the paper
Some of what he talks about:
"Soon enough you will be able to take any published research paper and tweak it, or improve it, any way you want"
"Why not write a program, or have an AI write it for you, that will take your favorite papers and improve them, and change their evaluations over time, as new results come in?"
"Will highly productive researchers, who publish a lot of papers, become far less valuable? The individual paper no longer seems scarce, or will not be in another year or two."
-
casualphysicsenjoyer
- Site Admin
- Posts: 12
- Joined: Sat Jun 20, 2026 8:35 am
Re: Society session ideas
We discussed as an idea, a program to feed papers into Lean and then have it verify mathematical claims. But this is hard for a bunch of technical reasons