News
Social science research - where being right less than half the time is fine
1+ hour, 41+ min ago (602+ words) Key topics: Social science studies show ~50% replication success rate SCORE project tests replicability, reproducibility, robustness Key issue: overconfidence in p Just less than half of your predictions turn out to be more or less right, about 10% are completely wrong, and…...
The coming reckoning: AI's rise and the chaos it leaves behind - Ivo Vegter
6+ mon, 2+ day ago (1752+ words) The meaning of "AI" has evolved from simple logic systems to marketing hype. Generative AI creates "AI slop," recycling and distorting human-made content. AI's rise threatens truth, creativity, and public discourse through automation. Truly disruptive technologies are relatively rare, despite…...
FT: Tim Harford - Why the maths used for decades in Economics needs a rethink
6+ mon, 2+ week ago (435+ words) Key topics: Economics relies heavily on complex, sometimes impractical mathematics. Critics argue traditional math is limited for real-world economic insight. Alte...
FT: Deep Mind and Open AI achieve gold at "coding Olympics" in AI milestone
6+ mon, 3+ week ago (374+ words) Key topics: Open AI GPT-5 and Deep Mind Gemini excel in ICPC "coding Olympics'AI models solved complex math and coding problems outperforming humans Breakthrough ma Former participants include Google co-founder Sergey Brin and Open AI's chief scientist Jakub Pachocki. The…...
From Thales to AI: Why existing knowledge still matters in scientific discovery
7+ mon, 3+ day ago (200+ words) Evolution from ancient to modern scientific thinking Risks of rushing innovation, especially with AI and space tech A summary of Elon Musk's research process: Develop a principle based on the evidence and assign a probability of truth to it Draw…...
The Economist: Fraudulent scientific papers are booming; a subset of journal editors is partly responsible
8+ mon, 1+ day ago (493+ words) Key topics: Scientific fraud is growing far faster than research output Paper mills and corrupt editors enable fake studies Pressure mounts on publishers to combat...