It’s mostly a marketing ploy to raise money and avoid regulation. Nobody really knows what it actually means. Avoid for now
Tags research (72)
Who’s hyping whom? We cover AI while AI puts our industry under pressure. Transformation has become the default framing. A special issue of Digital Journalism looks at the hype cycle.
Researchers found that chatbots, in their eagerness to please, are overly agreeable when giving interpersonal advice. GPT4o, Gemini-1.5-Flash, Claude Sonnet 3.7, and others are affirming users’ behavior even when harmful or illegal.
Anthropic asked 81,000 Claude users what they want from AI. People in lower-income countries are more optimistic than those in Europe or North America. Worry about job loss is the clearest predictor of negative sentiment. Qualitative research at this scale was hard to do before. Whether Anthropic should be the one doing it is a question the report doesn’t ask.
238k speeches: Guardian and UCL trained a non-generative ML model on 100 years of House of Commons debates. Both Labour and Conservative MPs are currently at or near their most hostile on immigration, driven by competition with Reform UK.
Just send the prompt twice? A new paper argues that repeating helps non-reasoning models. There’s a catch: The models tested (4o, Claude 3.7) are retired by now.
Is the image even real? Can we verify the facts?
Those questions framed the conversation at last Thursday's AI for Media Network gathering in Hamburg. 120+ representatives from media organizations and academia met to discuss AI in verification and research. It was the first time the event was hosted at SPIEGEL-Gruppe's Hamburg offices. Gerret von Nordheim, deputy head of SPIEGEL's fact-checking department, presented our in-house...
What is Claude? Anthropic doesn’t know either. Gideon Lewis-Kraus spent months inside the company, watching researchers try to understand their own AI. A good example of how to write about this technology without falling for the hype or waving it all away. (The New Yorker)
Can AI help crack the 1986 Olof Palme assassination? Anton Berg and Martin Johnson are using AI to reanalyze the huge archive of police material. Their podcast Spår follows the work (in Swedish).
Ask AI “what’s the biggest pay gap?” and it’ll miss negative numbers. Ask for “the company” with the top score and it’ll ignore ties. Paul Bradshaw tested ChatGPT, Claude, Gemini, and Copilot for data analysis – and catalogues where tools trip up. (Medium)
OpenAI co-founder Ilya Sutskever discusses the current state of generative AI, and talks about his new company SSI and what he plans to do with the three billion dollars he’s raised: develop learning AI systems. (Dwarkesh Podcast)
The Thinking Game: This documentary from 2024 takes us behind the scenes with Google DeepMind chief Demis Hassabis, showing how they taught their algorithms to play video games, chess, and Go, before moving on to protein folding. It’s compellingly told, though it sidesteps what was happening simultaneously at OpenAI and elsewhere. For free on YouTube.
Case Study: How the Financial Times Approaches Transparency about AI Use in News (Liz Lohn, Felix M. Simon)
