Homepage
Open in app
Sign in
Get started
Discovery at Nesta
Nesta’s Discovery Hub scans for emerging trends, promising technologies and interventions
Follow
Catalysing innovation for missions outside government
Catalysing innovation for missions outside government
The government’s five missions need to stretch beyond the public sector. They can only succeed by marshalling ingenuity from across…
Laurie Smith
Sep 9
Harnessing AI for qualitative interviews
Harnessing AI for qualitative interviews
Key insights from our evidence review
Natalie Lai
Aug 1
Planning an automated horizon scanning process
Planning an automated horizon scanning process
We introduce our upcoming work on an automated, data-driven horizon scanning approach
Tom Willcocks
May 6
Songs of change: data reveals how musicians advocate for the environment
Songs of change: data reveals how musicians advocate for the environment
How are musicians using their talent and voice to respond to the climate crisis?
Adeola O
Apr 25
Building an AI chatbot with Streamlit and OpenAI’s GPT-4 to drive engagement with online content
Building an AI chatbot with Streamlit and OpenAI’s GPT-4 to drive engagement with online content
How we used LLMs and prompt engineering to build a custom chatbot to engage readers with long-form content about future signals
Karlis Kanders
Feb 7
Using machine learning to discover innovations that can boost child development
Using machine learning to discover innovations that can boost child development
We will analyse data such as research papers and patents to spot impactful innovations for anticipating and managing child development…
Rosie Oxbury
Nov 7, 2023
How to evaluate large language model chatbots: experimenting with Streamlit and Prodigy
How to evaluate large language model chatbots: experimenting with Streamlit and Prodigy
We describe our initial experiments of using Streamlit and Amazon S3 to log user feedback, as well as performing A/B tests using Prodigy.
Karlis Kanders
Oct 17, 2023
How to build an early-years activity recommender with retrieval-augmented generation, GPT-4 and…
How to build an early-years activity recommender with retrieval-augmented generation, GPT-4 and…
We used GPT-4, Pinecone and Streamlit to generate early-years activities grounded in Development Matters best-practice guidance
Kostas Stathoulopoulos
Sep 4, 2023
How to use GPT-4 and OpenAI’s functions for text classification
How to use GPT-4 and OpenAI’s functions for text classification
A technical guide on how we mapped early-years activities to seven Areas of Learning from the Early Years Foundation Stages framework
Kostas Stathoulopoulos
Aug 31, 2023
Combining WhatsApp with large language models: prototyping with Twilio and Flask
Combining WhatsApp with large language models: prototyping with Twilio and Flask
Exploration of building a WhatsApp chatbot prototype to answer questions and generative personalised activities for early-years educators
Karlis Kanders
Aug 15, 2023
Using Streamlit and GPT-4 to prototype a web app for suggesting personalised early-years activities
Using Streamlit and GPT-4 to prototype a web app for suggesting personalised early-years activities
A technical guide on how OpenAI’s GPT-4 could assist caregivers by suggesting personalised activities in early-years education.
Kostas Stathoulopoulos
Aug 8, 2023
In praise of “boring innovation”
In praise of “boring innovation”
How dull can change the world
Laurie Smith
Jun 28, 2023
Can we harness generative AI for social good?
Can we harness generative AI for social good?
Nesta’s Discovery Hub is launching a new project to find impactful use cases of generative AI in education, health and sustainability
Karlis Kanders
Jun 16, 2023
Causal inference for discerning the impact of research grants on company success
Causal inference for discerning the impact of research grants on company success
We explore the use of causal inference methods to provide data-informed recommendations on policy and funding interventions.
Jack Rasala
Jan 17, 2023
Predicting start-up success using machine learning
Predicting start-up success using machine learning
In this blog, we experiment with using machine learning to predict which companies are more likely to go on to be successful.
Jack Rasala
Jan 17, 2023
Taming and training hunches: Data analytics and foresight
Taming and training hunches: Data analytics and foresight
How Nesta’s Discovery Hub harnesses data science to deepen our capacity to evaluate possible futures
Karlis Kanders
Jan 3, 2023
Discovery at Nesta: Experiments in strategic foresight
Discovery at Nesta: Experiments in strategic foresight
Welcome to the Medium blog for the Discovery Hub at Nesta.
Siobhan Chan
Jan 3, 2023
About Discovery at Nesta
Latest Stories
Archive
About Medium
Terms
Privacy
Teams