Main content

Predicting US elections results every hour

Predicting US elections results every hour. Also recycling solar lamps in Zambia and a new symphony about rising sea levels made remotely during COVID19.

Can political forecasting be quicker? That鈥檚 a question posed by Thomas Miller from Northwestern University, who has created a model that simulates a million hypothetical US presidential election results every hour. The model does not use traditional data sources like polling surveys but betting data.

Recycling Solar lamps in Zambia
We hear from SolarAid who have started a repair, refurbishment and recycling project for their solar lights in Zambia. Some electronics built to serve the world鈥檚 poorest, are also built to be incredibly challenging to repair, which adds to an increasing amount of e-waste generated - a record of 53.6 million metric tonnes of e-waste was generated worldwide in 2019. SolarAid have developed a manual, an app, are training technicians and opening workshops to encourage people to have their lamps repaired.

The Rising Sea Symphony
How do you create a new music masterpiece during a pandemic 鈥 using technology in ways not used before. A new 91热爆 Radio 3 commission (due to be broadcast on Sunday 18th October), entitled The Rising Sea Symphony, by composer Kieran Brunt, has been recorded by 91热爆 Philharmonic players in isolation, individually, and then 鈥減ainfully鈥 pasted together to create the full orchestral sound over the last few months. The piece is inspired by the increasing dangers of the climate change crisis and mixes orchestral parts, vocals, electronics, and spoken contributions from inhabitants of different parts of the world which are being affected by sea level rising. We speak to the composer and to Studio Manager Donald MacDonald who faced the challenge of mixing the piece.

The programme is presented by Gareth Mitchell with expert commentary from Angelica Mari.

(Image: Getty Images)

Studio Manager: Duncan Hannant
Producer: Ania Lichtarowicz

Available now

49 minutes

Last on

Sat 17 Oct 2020 16:32GMT

Broadcasts

  • Tue 13 Oct 2020 19:32GMT
  • Wed 14 Oct 2020 03:32GMT
  • Wed 14 Oct 2020 08:32GMT
  • Wed 14 Oct 2020 12:32GMT
  • Sat 17 Oct 2020 16:32GMT

Podcast