Iris.AI democratizes access to research
This company helps you to understand the science behind your favorite TED talks
You just saw a really cool TED talk on, say, lab-grown organs, and want to learn more about it, but you don’t know anything about biotechnology. Google Scholar would be a good place to start—if you knew the right keywords to search for. Enter Iris.AI.
Iris.AI can use the TED talk as input and find exactly what you’re looking for.
Iris.AI is a company that “aims to make scientific research fast and easy, even if you’re a total non-expert.”
The goal of Iris.AI is to build an artificial intelligence capable of highlighting new trends and interconnections of discoveries to help knowledge seekers connect the dots.
Even if the science behind your interest area is completely inscrutable to non-experts, the AI-trained methodology helps you search using all sorts of different inputs, so you can instantly access and navigate the research of your interest.
The company was first established at Singularity University in 2015, where founder Anita Schjøll Brede was one of the first Norwegians to complete the program. She now runs the Singularity University’s Global Impact Competition for Norway.
Millions of papers never even read
“Nearly 1.8 million research papers are published each year by researchers around the world. Yet, only a disturbingly small portion of it is put into practice or ever even read. These papers, and the connections they make, hold great potential for innovation. The challenge is that human brain is not powerful enough to find and process all the relevant scientific information that is available,” writes the company on their launch.
Though this sounds innocuous, the movement to democratize access to scientific research is remarkably controversial. The “Open Access” debate has been raging in scholarly circles for the past few years. Perhaps the most widely known embodiment of the movement was the arrest and subsequent suicide of Reddit co-founder Aaron Swartz, who was caught illegally downloading articles from the research database JSTOR.
Founder Anita Schjøll Brede feels that making scientific research more accessible is a simple matter of better knowledge sharing, which will be “immensely beneficial.”
“Our starting point was: scientific research is really hard to find. It is inaccessible. How do we solve that? We looked at all the different problems of research, the distribution, the way you interact, how you find it. … And what we found is that there are millions of papers out there, but you can’t find them. And the idea came along to use AI as a way to overcome this.”
Fear of AI
Open access isn’t the only controversial aspect of Iris.AI. AI has been a tech bogeyman since it first started coming into use, as people fear it will take over their jobs—or worse. “There’s a group of people that are scared of AI because they watched Terminator,” says Schjøll Brede.
But in the real world, AI companies have been focused on specific tasks—feeding your pets, scheduling meetings, and detecting languages. Which is exactly what Iris.AI is doing.
“We are building a very narrow AI that does one thing: she knows how to read science, and tell you about other science related to it.”
The World Economic Forum recently reported that AI is expected to deliver net job growth by 2020. Schjøll Brede is only positive about the vast potential for AI.
“It is not just about AI taking our jobs, but about saving our lives, making your life quality better, so, for example, that those billions of people who don’t have access to healthcare, can get access to healthcare.”
However, she acknowledges that it is important to talk about the issues surrounding AI, and the inevitable changes this technology will bring.
“So many people are using AI technologies for good, so yes, there will be a painful process where jobs change, but the world is getting radically better by the hour.”
This article was originally published on Nordic Startup Bits.
It also appeared in the Oct. 7, 2016, issue of The Norwegian American. To subscribe, visit SUBSCRIBE or call us at (206) 784-4617.