How Machine Learning Can Be Used for Good

Published on March 25th 2020
A group of people holding some plants
Machine learning has been a hot topic recently with polarising discussions emerging as to whether it’s ultimately good or bad for society.
The University of Massachusetts found that training one AI model can produce as much carbon as five cars.
Sasha Luccioni, a postdoctoral researcher at the Mila AI Institute in Quebec, worked on a tool to help AI researchers estimate the carbon footprint of their machine learning models.
She acknowledged that the University of Massachusetts paper was a unique case, because unlike the scenario used in the study, “very few people” train their models from scratch, and “a lot of training is now done using cloud services from companies such as Google, Amazon, and Microsoft, which are mostly carbon-neutral or moving toward it.”
While its important to acknowledge both sides of the argument, recognising the positive effects that AI can have on communities and society can inject some balance and perspective into the debate.
Researchers at The University of Illinois, for example, are working on tracking 10,000 whale sharks, an endangered species that play an important role in maintaining the health of our oceans. Their use of AI to collect and curate data will free up valuable time needed by researchers to develop and revise conservation plans to combat extinction.
In light of this, it’s worth remembering the purpose behind the technology. Does it improve the availability of knowledge for people? Does it draw awareness to hidden issues that might not have been otherwise known to people without using it?

Using technology to connect people with their passion

A girl using her phone in a green house
We hope that by using machine learning, we can make gardening and growing plants easier for people.
It’s important in the digital age that technology is immediately relevant to people’s lives. As a 2019 study conducted by researchers at the Technical University of Denmark found, people’s attention spans are narrowing due to the abundance of information available in the public domain.
This makes switching between tasks tempting and presents an extra challenge for digital platforms to make products that function in a useful, engaging way.
We are not ignorant of the potential every gardener has to increase biodiversity and reduce their carbon footprint by growing their own produce, so we are confident that by introducing technology that helps to engage people more, we are using technology for good.
Amine, our Data Scientist, is working on an algorithm that will surface more content that people are interested in based on the options they select when they first download Candide.
If someone says they are most interested in houseplants, we will show them more posts with them. If they express an interest in growing fruit and vegetables, then we will surface information to help them learn and progress their hobby.

How our technology is being used to bring people together

a woman taking a photo of a plant
Machine learning at Candide is present in many ways. The company is split up into four main areas; Marketplace, Knowledge, Socialising and Visting. Each of these teams uses machine learning or are planning to use it in the future to bring the community closer together.
We want to connect people to each other to share gardening tips, answer questions and generally socialise.
Through machine learning, we can learn more about the types of posts people have “dug” (liked) in the past. Using this information, we can then show people similar posts from others in the community, helping to support relationships and conversations between hobbyists.
In the future, we hope to use machine learning to surface posts of plants and gardening equipment relevant to the user's interest that they can buy from community members in Candide’s Marketplace.
If people post a lot about cacti in Candide, our aim is to show them any cacti available. They would then be able to message the seller to find out more, which helps create relationships and trigger knowledge sharing.
While our visiting team doesn’t use machine learning to directly influence our users, it’s been using it to showcase to gardens metrics like visitor number predictions, demographics and tracking busier and quieter times. This will have an impact on our community as events could be planned by gardens during predicted busier times, allowing people to socialise in beautiful spaces.
Our engineers are also developing a question and answer bot which has been created using supervised machine learning to train the machine to answer people’s gardening related questions.
We hope that this will engage people more in gardening and give them the tools to share the knowledge they’ve learnt with the community.
During a time where globally, people are relying on technology more to learn and connect with one another, we want to use our machine learning capabilities to encourage this. As research has shown, people who use digital platforms actively - by sending messages, liking tweets, sharing information - are happier.

Be the first to join our beta program

Help us build a place where community meets knowledge. Try it out and let us know what you think.
Download on the App StoreGet it on Google Play