We should use AI in the REF, and it is good to spend taxpayers' money on lizard venom research. Thanks to @resprofnews.bsky.social for inviting me to write up some thoughts on where Labour's new DSIT team should go on R&D, just in time for conference!

Strong argument from @sanjushdalmia.bsky.social for basic research funding "As a doctor, I meet patients whose diabetes management has been completely transformed by drugs such as the weight-loss medication Ozempic. This has only been possible thanks to curiosity-driven research on lizard venom"

Saw Kanishka Narayan speak at a Labour Digital thing a few months ago and he was incredibly impressive at knowing what he was talking about on tech stuff, which isn't always a given for politicians, so that seems like a good appointment!

🚨BREAKING🚨 Patrick Vallance announces three-month pause in REF. Science minister announces hiatus but says results for next Research Excellence Framework still on track for 2029. Free to read: www.researchprofessionalnews.com/rr-news-uk-r...

As the Labour government applies its interventionist industrial strategy to R&D, policymakers must carefully pick the right policy tools for the right problem. Yesterday, @britishprogress.bsky.social launched our interactive “R&D Policy Toolkit” to help policymakers do just that.

🚨🔧 NEW R&D POLICY TOOLKIT LAUNCHED 🪏🚨 Excited to launch a new @BritishProgress product: an interactive toolkit for UK R&D analysts and policymakers! 🔗 rd-toolkit.britishprogress.org Thread below🧵

"Repair" won't work - Starmerism needs to "Rebuild". Britain's stagnation is in part, down to the accumulation of decisions where Conservative governments ducked transition costs. But the political costs of inaction now exceed the political costs of transition.

“People don’t want to interact with a paper. They want to interact with the author” @valdavinson.bsky.social on lack of onsite commenting on papers at #FutureSciPub 1/2

I'm looking for academic affiliation. I don't need to be paid & I'm happy to offer something in return! I've got an exceptionally strong track record and full metascience research program ready to roll.

29 scholars reflect on their participation in adversarial collaborations: “Rather than producing a clear 'winner,' the most common outcome was a deeper understanding of the problem space through the integration of opposing perspectives” Open Access: doi.org/10.1007/s111... #MetaSci #Methodology 🧪

We can reform the Research Excellence Framework to distribute power, strengthen open access and unlock more money for R&D, while retaining both the “dual-support” system and qualitative research evaluation. In my latest Substack post, I outline what the “Open REF” could look like.

What counts as interdisciplinary research? And can we tell who is actually funding it, versus just talking about it? A data-driven deep-dive from me here: open.substack.com/pub/johnsonb...

What do we need to *build* for metascience and open science? I’ve written a “Request for Products” that I think could be particularly valuable to improve science. open.substack.com/pub/whitehea...

Inspired (partly) by ‪@jontreadway.bsky.social 's tweet thread, I have come up with a highly subjective and tentative map of different strands in metascience. It is intended as a starter to begin talking about what people (want to) include in metascience. #metascience2025

Today on my Substack, I'm sharing a list of empirical research questions for metascience, which I think are particularly relevant for science policy. I don’t conduct or fund metascience research, but if you do, you might find this list helpful. whiteheat.substack.com/p/empirical-...

I'll be at the Cambridge Centre for Science and Policy (CSaP) annual conference later today. Come say hi if you're attending - I'd love to chat about @britishprogress.bsky.social and Scientists for Labour!

The "for-profit company owned by a non-profit" model seems to have been extraordinarily successful for science (OpenAI, Novo Nordisk). This paper by @namratanarain.bsky.social shows VCs are short-termist - perhaps non-profits are better placed to support "patient science".

Lee Kuan Yew claimed air conditioning was "key to public efficiency". Yet we only have a few, low-quality studies on the productivity effects of office working environments (temperature, air filtration, noise etc). Potential big wins here, for both workers and businesses.

Last week, The Economist characterised Labour's political strategy as "all pain, no gain". Why was there so much anger on means-testing winter fuel payments, and how can Labour improve in the future? By renaming stuff: whiteheat.substack.com/p/rename-the... 🧵

🎯 Does early success in science funding shape the rest of your career? New RoRI study says yes. The ‘Matthew effect’ — early winners keep winning — holds across funders & countries. 📄 Read more: researchonresearch.org/largest-stud...

The next few months will determine the outlook for the UK #LifeSciences research landscape for a decade. An article from me on how Government needs to maximise the opportunities and avoid the pitfalls: news.cancerresearchuk.org/2025/06/05/5... 1/

Existing AI evals aren't that useful for guiding adoption and boosting productivity. We need more real-world trials run by the What Works Centres to inform and de-risk adoption, helping Labour to deliver on its Five Missions. Out today on my Substack: whiteheat.substack.com/p/ai-evals-f...

Metascience should complement evidence-informed policy by raising the quality of the *best available* evidence over time. Metascience should not be weaponised against evidence-informed policy just because evidence does not always meet extremely high standards of rigour.

I beg you please stop calling p-hacking exploratory analysis. Call it data dredging, fishing expedition, cherry picking, significance chasing. Labeling your non-pre-reg'd analyses clearly as exploratory isn't a license to p-hack. Stop bastardizing scientific exploration and stop hiding behind it.

There is a new Executive Order that is closely related to many of the topics promoted in the science reform movement. It deserves a close read as it will surely have implications for how some promoted reforms are implemented in policy and practice. My hot take 🧵 www.whitehouse.gov/presidential...

I am a complete doomer about cooptation, ftr. I more or less just think you should say and advocate for what you think is true, and can defend on good grounds, and never bother trying to hedge against people coopting your ideas. Ideas can always be coopted, there is no defence against that.

It's all so farcical. Nobody could seriously believe that there's some plan here which goes step 1: cancel all foreign student visas at Harvard, step 2: ?????, step 3: increase replication rates in biology. The lie is so transparent as to be insulting.

Thread relevant: Another form of weaponization of the reform movement is using reproducibility/credibility challenges as a justification for reducing investment in research. This is a logical error, like saying the faucet is leaking so the house should be burned down.

That is, scientific information could be disregarded entirely unless it meets an extremely high standard -- a gold standard! Then, what is decision-making based on? I testified that this would be counterproductive--"best available evidence" is important. My testimony: osf.io/preprints/me...

As people caricature what they think "metascience" is, I'd draw your attention to this piece of important metascience. It found that the vast majority of clinical prediction tools published during the pandemic were such poor quality they couldn't have possibly helped. www.bmj.com/content/378/...

I can’t seem to find surveys / qual research on the main self-reported reasons for *why* anti-immigration voters want less immigration. Anyone have links?

🇧🇷 The Brazilian Reproducibility Initiative published the results of 143 replication attempts in biomedical science. Success rates ranged from just 15–45%. Now, the project team reflects on what made replication so hard and what needs to change. buff.ly/55j9Sax

On my Substack today, I consider the impact of AI as a legibility-enhancing technology, and what this means for government intervention, using land value tax as an example. open.substack.com/pub/whitehea...

This Thursday 24 April from 6pm to 8pm, I’ll be joining scientists, policy wonks, journalists, Labour staffers, and possibly an MP or two at Scientists for Labour's next Science Policy Drinks event in Westminster. Link to sign up here: docs.google.com/forms/d/e/1F...

🚨 NEW PAPER KLAXON 🚨 @britishprogress.bsky.social's first research report is out! It is all about the Government's announcement of a new defence innovation unit – UK Defence Innovation. Here’s how to build this unit for maximum impact on security and prosperity: 🧵

I’m in London Tuesday April 22 and Wednesday April 23. Drop me a line if you’d like to get a coffee and chat about science and tech policy!

What are advance regulation and deregulation commitments? Drawing on the 2030 ban on petrol and diesel vehicles and a recent economics of innovation paper, I've written about two new tools to help deliver the government's industrial strategy and drive growth. open.substack.com/pub/whitehea...

Does anyone have a link to work identifying Britain's main export opportunities with regards to net-zero technologies? (other emerging technologies would be good too)

📣 New report 📣 In a volatile world, we need an economy that is firing on all cylinders - but regional imbalances are driving overheating and stagnation in different places. www.labourtogether.uk/all-reports/...

💫 We’re launching the Centre for British Progress Our founding essay: Rediscovering British Progress is a case for growth that drives shared progress, rooted in Britain's values and industrial heritage. It all starts with a postcard from 1870 👇 britishprogress.org/articles/red...

Excited to be joining the Centre for British Progress as a fellow today, building upon our work on science, innovation and technology policy at UK Day One.

I haven't seen that but you might like this one, which surfaces papers posted by people you follow

Fascinating findings: Of the *undergraduates* surveyed, 65% said that knowing what they know now they would make the same decision again. [...] 48% of the *graduates* surveyed said that with hindsight they would make the same decision and 52% a different decision. www.hepi.ac.uk/wp-content/u...

Re-reading the Danish intelligence report from February 2025, which, I think, is the least conservative assessment of Russian reconstitution timelines & capability among all the published views from European intelligence services in the past year. www.fe-ddis.dk/globalassets...

Any bluesky feeds mainly containing abstracts from econ papers / JMPs?

yup! Exactly. Even for the stated aim of saving bats, the £100 million is being badly spent there

An essay from @jo3hill.bsky.social which can be summarised as "Public interventions need clarity of outcome" reform.uk/publications...

Interesting/horrifying new report from UKHSA on infectious diseases. Notable increases in TB, gonorrhoea, syphilis, HIV, legionella, listeria, salmonella, measles, pertussis, and norovirus... Unsurprisingly, worse health outcomes in people living in more deprived areas. www.gov.uk/government/p...

We urgently need to scale up RCTs on AI in education. The RCTs we do have suggest impressive learning gains in some contexts (arxiv.org/abs/2402.09809) and harms in others (papers.ssrn.com/sol3/papers....). Gov should consider supporting the EEF with new $$$ and AI experts.