Tag: ai-research
32 articles tagged ai-research
Anthropic blames 'evil AI' fiction for past Claude blackmail attempts
Anthropic says training on internet text portraying AI as evil drove early agentic misalignment, and new alignment methods have cut blackmail rates from 96% to 0%.
Google Finance AI experience expands to over 100 countries
Google is rolling its AI-powered Finance experience out to more than 100 countries with local-language support, live earnings audio and expanded charting.
Alan Turing Institute Told to Reform After UKRI Review
UKRI says the UK's flagship AI research institute is underperforming on strategy and value for money, demanding a pivot to defence and national security.
UK Government Creates Fundamental AI Research Lab With £40m Backing
The government is establishing a new research lab to tackle AI hallucinations, unreliable memory, and unpredictable reasoning, backed by £40 million over six years plus significant compute access.
UK Commits Record £1.6 Billion to AI Research and Innovation Strategy
UKRI launches its first AI strategy backed by record funding, targeting breakthroughs in healthcare, clean energy and public services.
Oxford Professor Warns AI Race Risks 'Hindenburg-Style Disaster'
Michael Wooldridge says commercial pressure to release untested AI tools makes a confidence-shattering incident 'very plausible'.
AI Reconstructs Rules of Ancient Board Games Lost to History
The Digital Ludeme Project at Maastricht University uses AI and computational techniques from genetics to reconstruct the rules of board games played before 1875.
Oxford Study Finds AI Chatbots Give Inconsistent and Inaccurate Medical Advice
A University of Oxford study involving 1,300 people found that AI chatbots produce varied medical advice depending on how questions are worded, with users struggling to get useful guidance.
Perplexity Launches Model Council to Run Queries Across Multiple AI Models
Perplexity's new Model Council feature runs queries across three AI models simultaneously, with a synthesiser resolving conflicts and highlighting where models agree or disagree.
Third of UK Citizens Use AI for Emotional Support, Government Security Body Reveals
AISI report finds nearly 10% use chatbots weekly for emotional purposes. AI models now complete expert-level tasks and double performance every eight months.
Anthropic Reveals What 1,250 Professionals Really Think About Working with AI
New AI-powered interview tool uncovers how workers across industries are integrating AI, with scientists wanting partnership but struggling to trust it for core research.
AI use makes us overestimate our cognitive performance, study reveals
Study reveals AI users overestimate cognitive performance regardless of ability level. Higher AI literacy correlates with greater overconfidence.