Profile

I’m a scientist at the University of Nottingham. My job is to use computer intelligence to learn about atoms.

I got into science in primary school when I was 8. I wasn’t very smart, wasn’t very tall (I’m still not!) and didn’t really know what I liked or wanted to do. My teacher, Mr Mallet, bought me a book called “The Body Owners Handbook” and I just took to it. I don’t think I actually expected to be made to feel so happy to learn things that interested me. 15 years on, it has just moved with me to my first home, and was the first thing I wanted to pack.

My grandpa also really wanted to be a scientist, but back in those days not as many people could go to unversity, so he didn’t go. I found an amazing school textbook from 1944 that he bought after he left school,  just bought he wanted to find out more about the world. I like to think I do a PhD and do my research for him. It’s also super interesting because of just how old fashioned the science is. Some of the stuff he (and your teachers!) were taught is now known to be flat out wrong today.

I think it’s exciting that my work, stuff that’s seen as super fancy and high tech –  could become completely wrong and outdated in 80 years. Because that’s science, and that means that I contributed to human knowledge, and that’s why I love it. It’s your own part of the world, your own part of human history, and your own “legacy” that nobody can take away.

I love to play music – I play the cornet (a small trumpet), normal trumpet, flugelhorn, and a bit of piano. I play them in a brass band – we even get to go and compete with other bands and universities, too. (I’m also an amazing singer – just don’t ask my friends…).

Most importantly though, I’m a massive cricket fan – I was paid to go to Pakistan for a conference last year, and I managed to get the British Council to let us play cricket for a bit 🙂

I’m what’s called a “nanoscientist.” One of the things we like to do is look at atoms, such as Hydrogen, Silicon (which makes up your computers) and Gold. These are some of the building blocks of the entire universe, so it’s important that we understand what they look like and how they interact with each other. Sometimes we just want to look at individual ones, but the majority of the time we want to make new molecules with specific properties. We also don’t want to have to build them all by hand, so understanding how they interact at a simple level is super important, because we can then make them build themselves!

Because atoms are tiny (Make a 1 millimetre gap in your hands… now make it 10 times smaller, now do it again six times!), we need super fancy technology to be able to “see” them. The microscopes that biologists use are about 1 million times less powerful than ours. So instead of using visible light to see them, my team has several of these things called “scanning probe microscopes.” Not only do they let us see individual atoms, but they even let us move them about.

https://www.youtube.com/watch?v=oSCX78-8-q0

It’s surprisingly simple, too.  My boss taught his 8 year old daughter to do it!

These machines are absolutely massive, and cost over £1 million each! I’m super lucky, because in Nottingham we have more than 4 of these for my research group for us to share. Here’s one of them – I really want to get back to use these again!

So why am I in the coding zone?

Well, like you, computers are very good at following exact instructions (take a mug, put a teabag in it, add boiling water, wait 120 seconds, add 10ml of milk, add 1 teaspoon of sugar, thank you for the cuppa!). But unlike us, they don’t have any common sense! So if we ask them to do something harder (is this a cat? is this a dog? is this bit of skin a cat’s whisker or a piece of fur?), we can’t turn it into exact instructions 🙁

This is the problem we have when we look at atoms. Because everything we do is so teeny tiny, a tiny change in our sensor completely changes what we see! We may go from seeing circles, to blurry circles, squiggles, ovals, and so on (this surface is silicon, and is MEANT to look like the picture in (a) and (b). Except as you can see in the other photos, it doesn’t.

These are things we can’t separate apart with an exact list of instructions – so even though it’s easy for a primary school student to see the shapes, a computer can’t! Even worse, not only do we have to babysit the equipment in case this happens, but we have to spend nearly half of every day fixing it!

So I’m responsible (along with several other people in the world) for trying to teach computers to help do it for us. People have been trying to do this since the 1990s, and it’s only now we have figured out ways to make real progress doing it! At the moment my code tells us, live, if it’s working properly or not. The future plan is to turn fixing it into a game, and then teach AIs to start fixing it for us!

This sounds fancy, but the truth is that we think “hey, this might work,” and try something and see. There isn’t a fancy method, there isn’t a big robot, there isn’t a huge amount of smarts required. My job is basically to just try stuff until I think it works, then tell other scientists what works and what doesn’t.

Having said that, I am super lucky because you need computers designed for gaming to do this learning, so I have no choice but to use a really fancy computer (this is 1 of the 2 I get to use).

One of the things we are working in lockdown is using 100,000s of nanoscale simulations to train AIs to help us process and search through the terabytes of data we generate over the years, without ever having to spend time searching through hours and hours and hours of data to find the one specific thing we’re looking for. The idea of combining simulations and real-world data never gets old…

The University of Nottingham is a public park, so every day I get walk to go to work through a forest and along a lake. It wakes me up, and gets me ready for a day of thinking about science. Here’s me walking to my department on a good day last year, with my final university project:

(I’m jumping here, by the way. I’m not standing on the water, as much as I wish I could.)

When I get to my office, the first thing I normally do is look at the data our computers calculated overnight (unlike us, computers don’t need to sleep, and don’t complain about long hours). We figure out and write down what has happened, and if we think what we did helped the computer to learn better or not, look into that a bit more. We then come up with some things to try for the next day – these have to be ready before I can go home!

I normally then go to the people working in the lab and discuss what I am doing with them. One of the most important things is to ask each other questions! No scientist can do it by themselves – they need friends to ask for help and ideas! They also know far more than me about how our machines operate – there’s no point in making fancy computer code that will be worthless in the real world! After collating some ideas, we all stop working at 11am to have a break, so we can do just that! My research group also all meet up every week or two and tell each other what we are all doing, so we can share ideas better. I bring cookies if I have a lot to say!

So I then spend my day asking questions, answering emails from other scientists in the world who we get to work with (for me, it’s Canada and the Netherlands at the moment), writing computer code and testing my ideas. Once everything is ready to run, I’m normally done for a day. I finish what I’m doing, and walk with my friends back along the lake and home, where I can rest up for another day of work!

I also do other things I do to help out the school of physics, like going to primary schools to talk about physics and showing new students doing A Levels exactly why they should come and come and study physics.