Timothy O'Leary
@timothyoleary.bsky.social
Professor of Information Engineering and Neuroscience, University of Cambridge
Hmmm. If science isn't already machine readable then I blame the machines...
"Science should be machine-readable" www.biorxiv.org/content/10.6... "This work uses Claude and was partially funded by Anthropic through their AI for Science program."
February 6, 2026 at 6:12 PM
Hmmm. If science isn't already machine readable then I blame the machines...
UKRI will more than double its AI research budget to £397m. The R&D+compute budget of OpenAI = $8.5b.
A single company has nearly 20 times the research budget of an entire G7 country. The lesson is simple: we can't chase the same research goals. We need to take risks that VC-backed companies won't.
A single company has nearly 20 times the research budget of an entire G7 country. The lesson is simple: we can't chase the same research goals. We need to take risks that VC-backed companies won't.
January 30, 2026 at 10:10 AM
UKRI will more than double its AI research budget to £397m. The R&D+compute budget of OpenAI = $8.5b.
A single company has nearly 20 times the research budget of an entire G7 country. The lesson is simple: we can't chase the same research goals. We need to take risks that VC-backed companies won't.
A single company has nearly 20 times the research budget of an entire G7 country. The lesson is simple: we can't chase the same research goals. We need to take risks that VC-backed companies won't.
Reposted by Timothy O'Leary
Moderna have announced that they won't run new phase three trials now with massive impacts for new vaccine development- why- its because of RFK Jr and his anti-vaccine campaigns and cancelling mRNA vaccine research which affect the sales the company can then make 🧪🧵 #PublicHealth
Moderna Won’t Run Phase III Vaccine Trials as Skepticism Grows in US: Bloomberg
Growing opposition to vaccines in the U.S., driven by recent government policy changes, makes it difficult to see a return on investment in vaccine development, Moderna CEO Stéphane Bancel said this w...
www.biospace.com
January 27, 2026 at 8:50 AM
Moderna have announced that they won't run new phase three trials now with massive impacts for new vaccine development- why- its because of RFK Jr and his anti-vaccine campaigns and cancelling mRNA vaccine research which affect the sales the company can then make 🧪🧵 #PublicHealth
Reposted by Timothy O'Leary
If you have tenure consider publishing less... doi.org/10.1038/d415...
I’m going to halve my publication output. You should consider slow science, too
If we don’t slow down, the research enterprise is going to crash, argues Adrian Barnett.
doi.org
January 19, 2026 at 9:59 PM
If you have tenure consider publishing less... doi.org/10.1038/d415...
I worry that wikipedia is losing ground to LLMs, like google gemini. Should I worry? Is this happening? Is this a bad thing?
wikipedia turns 25 today! the last unenshittified major website! backbone of online info! triumph of humanity! powered by urge of unpaid randos to correct each other! somehow mostly reliable! "good thing wikipedia works in practice, because it sure doesn't work in theory" - old wiki adage
January 15, 2026 at 3:31 PM
I worry that wikipedia is losing ground to LLMs, like google gemini. Should I worry? Is this happening? Is this a bad thing?
Reposted by Timothy O'Leary
when doing neuroscience projects I often advocate for computational modelling, followed by data analysis to test model's predictions.
however a few times now I have had pushback from collaborators/reivewers suggesting it would be better to do the data analysis first, then the modelling.
thoughts?
however a few times now I have had pushback from collaborators/reivewers suggesting it would be better to do the data analysis first, then the modelling.
thoughts?
January 12, 2026 at 12:47 PM
when doing neuroscience projects I often advocate for computational modelling, followed by data analysis to test model's predictions.
however a few times now I have had pushback from collaborators/reivewers suggesting it would be better to do the data analysis first, then the modelling.
thoughts?
however a few times now I have had pushback from collaborators/reivewers suggesting it would be better to do the data analysis first, then the modelling.
thoughts?
Or is this just evidence that the Royal Society lost relevance a long time ago? Having this tool on their books at least gives them the occasional headline.
If the @royalsociety.org won’t stand up for scientific values like truth, evidence & the importance of not spreading misinformation on an industrial scale, then who will?
www.theguardian.com/science/2026...
www.theguardian.com/science/2026...
Royal Society president reignites Elon Musk row by defending lack of action
Society should only eject fellows for fraud or other defects in their research, says Paul Nurse
www.theguardian.com
January 14, 2026 at 9:40 AM
Or is this just evidence that the Royal Society lost relevance a long time ago? Having this tool on their books at least gives them the occasional headline.
Reposted by Timothy O'Leary
Would love to hear expert views on this paper. It appears to show that the operationalization of brain activity the field has relied on for 3 decades—the BOLD response—is not actually a sensible measure of brain activity.
www.nature.com/articles/s41...
www.nature.com/articles/s41...
BOLD signal changes can oppose oxygen metabolism across the human cortex - Nature Neuroscience
Using quantitative brain imaging, the authors show opposite fMRI BOLD signal to metabolic activity due to variable oxygen extraction across the human cortex. This questions the canonical interpretatio...
www.nature.com
January 5, 2026 at 11:55 AM
Would love to hear expert views on this paper. It appears to show that the operationalization of brain activity the field has relied on for 3 decades—the BOLD response—is not actually a sensible measure of brain activity.
www.nature.com/articles/s41...
www.nature.com/articles/s41...
Reposted by Timothy O'Leary
What is the computational role of dendritic excitations? Byung Hun Lee and team mapped voltage dynamics throughout the dendritic trees of CA1 pyramidal neurons in mice navigating in virtual reality. www.biorxiv.org/content/10.6...
January 4, 2026 at 6:17 AM
What is the computational role of dendritic excitations? Byung Hun Lee and team mapped voltage dynamics throughout the dendritic trees of CA1 pyramidal neurons in mice navigating in virtual reality. www.biorxiv.org/content/10.6...
January 2, 2026 at 4:10 PM
Reposted by Timothy O'Leary
I don't always get pretty, isolated, AND transfected primary neurons in culture, but when I do I take advantage.... Rat hippocampal neuron overexpressing ThymosinB4-mScarlet and imaged for 16hr on a @zeiss-microscopy.bsky.social LSM880 with Airyscan. #FluorescenceFriday #Microscopy
December 19, 2025 at 2:01 PM
I don't always get pretty, isolated, AND transfected primary neurons in culture, but when I do I take advantage.... Rat hippocampal neuron overexpressing ThymosinB4-mScarlet and imaged for 16hr on a @zeiss-microscopy.bsky.social LSM880 with Airyscan. #FluorescenceFriday #Microscopy
I had a PhD application with an entirely hallucinated reference section. When I asked the kid why I couldn't find the journals let alone the references, he doubled down and said they were placeholders for the actual papers and the rest was definitely his work. We can't keep on top of this.
Grading and googling hallucinated citations, as one does nowadays, and now that LLMs have been around for a while, I've discovered new horrors: hallucinated journals are now appearing in Google Scholar with dozens of citations bc so many people are citing these fake things
December 16, 2025 at 10:43 AM
I had a PhD application with an entirely hallucinated reference section. When I asked the kid why I couldn't find the journals let alone the references, he doubled down and said they were placeholders for the actual papers and the rest was definitely his work. We can't keep on top of this.
Reposted by Timothy O'Leary
When your PI presents your work
December 15, 2025 at 12:00 PM
When your PI presents your work
Researchers extract latent representation of humanity's future from latest AI model
December 12, 2025 at 9:59 PM
Researchers extract latent representation of humanity's future from latest AI model
"a control system"
The "what else" argument is really something quite bizarre. The brain is a computer, because what else, a magic box?
Like a cat must be a horse, because what else, an elephant?
Like a cat must be a horse, because what else, an elephant?
December 11, 2025 at 12:35 PM
"a control system"
Meanwhile, applicants explore use of AI in writing proposals and... fuck it, everything.
UKRI exploring use of AI in grant review - this will prompt debate but interested to know if people think it’d encourage novelty or the opposite? www.chemistryworld.com/news/ukri-op...
UKRI opens up grant proposal data to explore using AI to smooth peer review
Surging applications has led the UK's main funder to look at ways to reduce the burden on reviewers
www.chemistryworld.com
December 4, 2025 at 9:56 PM
Meanwhile, applicants explore use of AI in writing proposals and... fuck it, everything.
With the US demonstrating its role as the paragon of democracy, unsullied by bribery and manipulation by wealthy individuals, the UK is desperate to follow.
giftarticle.ft.com/giftarticle/... Reform UK receives record £9mn donation from Christopher Harborne
giftarticle.ft.com/giftarticle/... Reform UK receives record £9mn donation from Christopher Harborne
Reform UK receives record £9mn donation from Christopher Harborne
Nigel Farage’s party attracts far more funding than Labour and the Conservatives
giftarticle.ft.com
December 4, 2025 at 9:30 PM
With the US demonstrating its role as the paragon of democracy, unsullied by bribery and manipulation by wealthy individuals, the UK is desperate to follow.
giftarticle.ft.com/giftarticle/... Reform UK receives record £9mn donation from Christopher Harborne
giftarticle.ft.com/giftarticle/... Reform UK receives record £9mn donation from Christopher Harborne
Reposted by Timothy O'Leary
Artificial intelligence promises rapid and polite feedback on papers — but we must first review the reviewer. This stand first says it all…
A timely World View Colin in our pages by Giorgio Gilestro
🧪 #academicSky
@nature.com
www.nature.com/articles/d41...
A timely World View Colin in our pages by Giorgio Gilestro
🧪 #academicSky
@nature.com
www.nature.com/articles/d41...
AI reviewers are here — we are not ready
Artificial intelligence promises rapid and polite feedback on papers — but we must first review the reviewer.
www.nature.com
December 3, 2025 at 2:09 PM
Artificial intelligence promises rapid and polite feedback on papers — but we must first review the reviewer. This stand first says it all…
A timely World View Colin in our pages by Giorgio Gilestro
🧪 #academicSky
@nature.com
www.nature.com/articles/d41...
A timely World View Colin in our pages by Giorgio Gilestro
🧪 #academicSky
@nature.com
www.nature.com/articles/d41...
This kind of approach is crucial for a theory of intelligence:
1/ Why does RL struggle with social dilemmas? How can we ensure that AI learns to cooperate rather than compete?
Introducing our new framework: MUPI (Embedded Universal Predictive Intelligence) which provides a theoretical basis for new cooperative solutions in RL.
Preprint🧵👇
(Paper link below.)
Introducing our new framework: MUPI (Embedded Universal Predictive Intelligence) which provides a theoretical basis for new cooperative solutions in RL.
Preprint🧵👇
(Paper link below.)
December 3, 2025 at 8:21 PM
This kind of approach is crucial for a theory of intelligence:
We are opening a FACULTY POSITION (tenure track, permanent) in the University of Cambridge at the interface of control and biology, interpreted broadly. Theorists and wet lab quantitative biologists with backgrounds in control, EE, applied math, ... apply by Jan 28!
www.cam.ac.uk/jobs/univers...
www.cam.ac.uk/jobs/univers...
University Assistant/Associate Professor in Control Theory and Systems Biology
Applications are invited for a University Assistant/Associate Professorship in the broad area of Control Theory and Systems Biology. The successful candidate will join the Control Group
www.cam.ac.uk
December 3, 2025 at 12:19 PM
We are opening a FACULTY POSITION (tenure track, permanent) in the University of Cambridge at the interface of control and biology, interpreted broadly. Theorists and wet lab quantitative biologists with backgrounds in control, EE, applied math, ... apply by Jan 28!
www.cam.ac.uk/jobs/univers...
www.cam.ac.uk/jobs/univers...
Reposted by Timothy O'Leary
The best scientific papers are provocations, not results you can rely on. Discuss!
What I mean is that they should try to force progress by making an outrageous statement that the established field wants to be wrong, but do it so well that proving it wrong is a real challenge.
What I mean is that they should try to force progress by making an outrageous statement that the established field wants to be wrong, but do it so well that proving it wrong is a real challenge.
November 30, 2025 at 12:59 AM
The best scientific papers are provocations, not results you can rely on. Discuss!
What I mean is that they should try to force progress by making an outrageous statement that the established field wants to be wrong, but do it so well that proving it wrong is a real challenge.
What I mean is that they should try to force progress by making an outrageous statement that the established field wants to be wrong, but do it so well that proving it wrong is a real challenge.
Hmm.... ya think?
I'm worried that there are some things slipping through (or soon to) as: LLM-generated manuscript, goes to editor using LLM to summarize, goes to reviewers using LLM to review, and not a single bit of the whole process is done by a human, is actual science, or is useful at all.
November 28, 2025 at 5:18 PM
Hmm.... ya think?
Reposted by Timothy O'Leary
Springer-Nature statement
“Whilst the details of peer review are confidential, we can confirm that the article underwent two rounds of review from two independent peer reviewers, supporting an accept decision.”
How am I now expected to believe that two people looked at the paper twice and DGAF?
“Whilst the details of peer review are confidential, we can confirm that the article underwent two rounds of review from two independent peer reviewers, supporting an accept decision.”
How am I now expected to believe that two people looked at the paper twice and DGAF?
Riding the Autism Bicycle to Retraction Town
Does anyone *really* know their Factor Fexcectorn?
nobreakthroughs.substack.com
November 28, 2025 at 1:36 PM
Springer-Nature statement
“Whilst the details of peer review are confidential, we can confirm that the article underwent two rounds of review from two independent peer reviewers, supporting an accept decision.”
How am I now expected to believe that two people looked at the paper twice and DGAF?
“Whilst the details of peer review are confidential, we can confirm that the article underwent two rounds of review from two independent peer reviewers, supporting an accept decision.”
How am I now expected to believe that two people looked at the paper twice and DGAF?