How smart are state of the art large language models? Diego Miranda-Saavedra weighs up what they can and can’t do and how we might position them with respect to our own capabilities #AI #LLMs realworlddatascience.net/ideas/posts/20…

On AI ethics - influencing its use in the delivery of public good - a new piece from our colleague Olivia Varley-Winter for Real World Data Science realworlddatascience.net/ideas/posts/20…

Data, data, data is often the mantra for robust AI model development, but data quality is often more important than data quantity and can throw up some interesting challenges Discussed in the latest in our AI series ⬇️ realworlddatascience.net/ideas/posts/20…

The black boxes often with unknown biases. AI ethics can be a minefield. In the latest in our AI series, @Ovarwin shares some recent insights and describes how everyone has a role to play realworlddatascience.net/ideas/posts/20… Royal Statistical Society @AmStatNews

The real world can be a dangerous place for an AI algorithm, full of unforeseen data, evolving behaviours and significant human consequences Dr Isabel Sassooon highlights some of the hurdles to navigate when assessing whether a model is fit for purpose realworlddatascience.net/ideas/posts/20…

"The only thing I am sure of is that there is no way of knowing how many jobs will be replaced by generative AI” (Oxford University's Carl Benedikt Frey) US pilot scheme looks at how to tackle the dearth of useful data navigating AI's workforce impact realworlddatascience.net/ideas/posts/20…

“What is it that you want to solve? What are you trying to do? Is AI the right thing? Can you do it with the resources you have available? That's a fundamental starting point.” – Janet Bastiman, the round table discussion on #AI Best Practice realworlddatascience.net/ideas/posts/20…

Optimising AI for direct users and applications can be knotty enough but how do you optimise AI developments for society as a whole? In the last of this article series on #AI we look at the governance strategies that can tackle the problems emerging now realworlddatascience.net/ideas/posts/20…

From banks tracking the impact of fiscal policies to humanitarian agencies monitoring displaced populations, "nowcasting" can be very useful Now researchers demonstrate how to make nowcasting estimates adaptable and more accurate realworlddatascience.net/ideas/datascie…

With escalating fears of online harms, real-world inequalities seeping across to social media networks & lack of access to the internet disproportionately affecting women across the world – are we muting the female voice online?#DigitalGenderGap realworlddatascience.net/ideas/posts/20…

Data gaps are hampering responses to real-world challenges. The US Census Bureau is bringing together sources – from structured numeric data to unstructured text documents scraped from the internet – to create statistical products that better serve society realworlddatascience.net/case-studies/p…

Finding the common ground in the apparently different #MachineLearning algorithms that won the #NobelPrize2024 awards for Physics and Chemistry may help us figure out how today’s powerful “black box” algorithms do what they do realworlddatascience.net/ideas/posts/20…

What steps do you need to follow to create useful statistical products from different data types? Learn from the US Census Bureau’s new approach to addressing real-world problems realworlddatascience.net/case-studies/p…

A skilled nursing facility use case demonstrates how effectively combining data from different sources using a new framework from the US Census Bureau can improve organisational resilience realworlddatascience.net/case-studies/p…

What next for the Curated Data Enterprise? The US Census Bureau wants to understand users’ experiences of the framework and is looking for feedback from research projects guided by the framework realworlddatascience.net/case-studies/p…

📣 RWDS is back 📣 You may have noticed our brief hiatus. Now we’re back - ready to rejoin the conversation with renewed clarity and purpose. Read the RWDS Editorial Board's new editor's blog and check out our Call for Submissions ➡️ realworlddatascience.net/the-pulse/edit…

The new addition to the RSS's suite of journals, RSS: Data Science and Artificial Intelligence, has published its inaugural editorial⚡⚡ We're looking forward to working hand-in-hand to strengthen the link between research and application 🤝 academic.oup.com/rssdat/article…

"I think of it like the medical version of the ghost of Christmas future but using a chain of events rather than clinking ghost chains." 👻 Will Browne argues for a UK sovereign AI model trained on NHS data. realworlddatascience.net/the-pulse/post…

Keeping abreast of guidance for industry decision-makers is crucial for data practitioners seeking to foster collaboration with their non-technical stakeholders. We review 'AI for Business' (BCS, The Chartered Institute for IT ) on The Pulse realworlddatascience.net/the-pulse/post…

"Like releasing a highly motivated intern into a complex archive with partial instructions, limited supervision, and the freedom to decide which filing cabinets, databases, or tools to open next". Our new case study evaluates Agentic AI 🤖 bit.ly/3UpQvMb