The top three news stories of the week, as chosen by our resident students. This week’s top stories include seeing stars, the brain rewarding us when we do good thinking, and robots in a maze!

By Fuad Mosis

Story 1

A long time ago in a galaxy far, far away…

In a distant galaxy, some of the earliest stars created since the beginning of the universe have been detected. Finding this galaxy, MACS1149-JD1, stretched the limits of modern technology and has shown how we can discern stellar events as early as a few hundred million years after the big bang. The stars were observed as they would have looked 550 million years after the big bang, some of which were 300 million years old at this point. This means these stars were born when the universe was 250 million years old, and provides some of the earliest insight into the ages of the early universe. Have a read of the original research article here.

Eureka!

Have you ever spent a long time trying to figure something out, and then it hits you? The chemical nature of this Aha!-moment has now been found. 30 participants were tested by being given word puzzles while simultaneously having their brains scanned. Researchers found that a dopamine release is triggered at the moment when a participant figured out the puzzle, contributing towards the positive reinforcement of completing the challenge. Dopamine is a chemical in the brain that can affect mood and emotions. This arguably makes sense considering that problem-solving is key to our survival as individuals, particular during our formative years as a species. Check out the original article here.

Story 2

AI take another ‘step’ forward

Grid cells in the brain are important for us to navigate and help us be aware of our spatial position – in other words they’re like an internal GPS. AI researchers from UCL and Google DeepMind created digital grid cells and tested how they affect the performance of an AI navigating a simple maze. They saw that the presence of these digital grid cells allowed the AI to reach the end of the maze more efficiently by taking shortcuts that were otherwise ignored. This study helps to elucidate both how grid cells work while opening new doors in AI research. Here is the original article.

Story 3