Academia has a new preprints problem. Thousands of researchers are essentially playing slot machines with large language models, pulling the lever over and over hoping to hit the jackpot of novel theoretical physics insights. As Figshare acts as a preprint platform, I'm watching this unfold in real-time. It&

Academic Publication Explosion Model Gradual AI Adoption Rapid AI Takeover AI Explosion Custom Scenario AI Adoption Year: Growth Multiplier: 10x Adoption Speed: 5 Model Assumptions: This model assumes AI adoption follows a sigmoid curve, with publications growing exponentially once AI tools become mainstream. The peer review crisis threshold is estimated

At SXSW London, I had the pleasure of seeing DeepMind co-founder Demis Hassabis speak on the future of artificial intelligence. Among many thought-provoking points, two remarks stuck with me. First, he emphasised the importance of understanding the fundamentals. Second, he championed the scientific method as a guiding principle for making

A few years ago, I wrote that Open Access is an inevitability. And in many ways, the data supports that view—at least at a glance. Gold Open Access—the model where authors (or their funders) pay article processing charges to make work freely available—has grown steadily over the

The 'Make Data Count Data Citation Corpus' looks for both DataCite DOIs and Accession numbers in the published literature. It basically tells you that "this publication DOI" mentions "this DataCite DOI or Accession number". By taking this unstructured data and combining it with the

The landscape of grant funding in North America has recently, rapidly, evolved, leading to a sense of nervousness and even pessimism within academia. However, remarkable technological advancements emerging from foundations and companies are counteracting this trend and fostering a renewed sense of optimism.  In the last week, we have seen

AI agents have the potential to revolutionise how researchers approach complex problems and manage information in academic research. These intelligent systems can automate tasks, analyse vast datasets, and even generate hypotheses, significantly accelerating the pace of discovery and innovation. AI agents can be categorised into different types, based on their