The Powerhouse team has launched a new initiative called “Do AI Differently” that calls for a human-centered approach to future development.
For years, we treated the output of AI like the result of a huge math problem. However, researchers behind the project from the Alan Turing Institute, the University of Edinburgh, AHRC-UKRI and the Registered Foundation of Lloyd say it’s wrong.
What AI is creating is essentially cultural artifacts. They are more like novels or paintings than spreadsheets. The problem is that AI creates this “culture” without understanding it. It’s like someone who memorizes a dictionary but doesn’t know how to have a real conversation.
This is why AI often fails when “nuance and context are the most important.” The system does not have the “depth of interpretation” to get what you’re actually saying.
However, most of the world’s AI is built on a handful of similar designs. The report calls this the “homogenization problem” and must overcome future AI development.
Imagine that every baker in the world used the exact same recipe. You’ll get a lot of the same, and frankly, boring cakes. With AI, this is copied with the same blind spot, same bias, same limitations and pasted into thousands of tools used daily.
I saw this happen on social media. It unfolds with simple goals and we now live with unintended social consequences. “Do AI Differently” teams are turning alarms to avoid making the same mistakes as AI.
The team has plans to build a new kind of AI. It’s about designing the system from the start and working as people do. Ambiguity, multiple perspectives, and deep understanding of context.
Vision is to create interpretive techniques that can provide multiple effective perspectives rather than one rigorous answer. It also means exploring alternative AI architectures to break the current design type. Most importantly, the future is not AI will replace us. It’s about creating an ensemble of human beings that we work with to combine our creativity with AI processing power to solve major challenges.
This can be exposed to our lives in a very realistic way. In healthcare, for example, experiences with doctors are not just a list of symptoms, they are stories. Interpretative AI helps you capture that complete story and improve your trust in your care and systems.
For climate action, it can help bridge the gap between global climate data and the unique cultural and political realities of local communities, creating solutions that actually work on the ground.
A new international funding call has been launched to connect UK and Canadian researchers on this mission. But we are at the intersection.
“We are in a crucial moment for AI,” warns Professor Hemiment. “We have a narrow window to build interpretive capabilities from scratch.”
For partners like the Lloyd’s Register Foundation, it all comes down to one thing. It’s safe.
“As a global safety charity, our priority is to ensure that future AI systems are deployed in a safe and reliable way, regardless of their shape,” says their technical director, Jan Przydatek.
This is not just about building better technology. It is about creating AI that helps us solve our biggest challenges and amplifying the best parts of our own humanity in the process.
(Photo by Ben Sweet)
reference: AI obsessions sacrifice human skills to us
Want to learn more about AI and big data from industry leaders? Check out the AI & Big Data Expo in Amsterdam, California and London. The comprehensive event will be held in collaboration with other major events, including the Intelligent Automation Conference, Blockx, Digital Transformation Week, and Cyber Security & Cloud Expo.
Check out other upcoming Enterprise Technology events and webinars with TechForge here.
