The results?

You'll have to go and find out!

I will say it has given me some optimism around the role of AI as *one* tool in science for overcoming bottleneck tasks that are graced with good data but large combinatorial solution spaces to navigate, even if I remain sceptical of it in other ways.

We collect over 5 million publications, clinical articles, patents and protein structures that build on AF2 and other high impact structural biology papers published since 2018.

We track whether AF2 adoption is associated with higher productivity, more novel protein structures and applied research.

Fwiw I dont think giving researchers credit for ideas in the form of citation counts is actually a great idea. The is more of a provocation to help us think about how we make the dark matter visible, rather than have it languish in funders' archives.

Shame you'll be too early for this

Hey Nicole. Congratulations. This sounds great! I've sent you an email. Myself and colleagues would love to have a chat about the metascience angle.

Anyway, sign up, give it a go and let me know what you think through the feedback button.

It's only going to be live for a few weeks ⏳

academic-idea-exchange.replit.app

Well... kind of. I've jumped on the vibe coding bandwagon to bring you this prototype app "This One for Free" - a platform to share and discover research ideas.

Think of it as more of a metascience provocation in the name of making the 'dark matter' of hidden research ideas more visible.

That's fair. I don't think the biggest problem is really individuals voicing grievances (shame the article focuses on that). But extensions like the valourisation of lived experience play out at movement level, are held up by a wide range of people, and narrow what is seen as possible or desirable.

But the latter feels significant to me. Liberal identity politics can be clumsily used on the left, diverting useful energy, hindering more collective politics, and creating unhealthy spaces and discourse. Not an issue in every left/progressive space, but most of us will have experienced it.

I think reading these arguments as "this is why the right is winning" vs "this is why the left isn't" matters.

The former seems weaker - 'woke' politics might push a small % people rightwards, but the right weaponises anything that can be used to divide. If it wasn't this it'd be something else.

Tbh gotta respect the ideological commitment compared to the current dudes

My bet is that most productivity gains will be where specific AIs contribute to producing tangible products by helping design better and more efficient materials, industrial processes and functionality. Not convinced gains from LLMs on knowledge work and services will be revolutionary or net useful.

The actual liberal blindspot is thinking the only 2 options are to endorse positions that get fascists rubbing their hands with glee or to write off entire segments of the population as irredeemable