It was 1995, the year that Craigslist, eBay and Expedia were born.

The age of the internet had arrived, and we at Forbesmagazine, all too aware of academics’ complaints about cashing out for research, made a prediction: Elsevier, the largest publisher of scientific journals, would be its “first victim”.

Yet recent years have seen Elsevier profits swell to more than £900 million closing in on a 40% profit margin. It seems to be—as the Financial Times claimed in 2015—“the business the internet could not kill”.

This hasn’t stopped resentment from brewing as journal prices continue to rise above inflation.

In the U.K. universities paid an average of nearly £4 million ($5.4 million) for journal subscriptions last year, with the majority of this budget going to five academic publishing companies: Elsevier, Springer, Wiley-Blackwell, Taylor & Francis and Sage.

At the close of 2017, a group of 200 German Universities even stood together against Elsevier, sparking further unrest in Sweden alongside renewed outrage from European academia.

Blockchain to the rescue

Could blockchain be what brings science’s untouchable publishing giants down? Norway’s Anita Schjøll Brede, cofounder and CEO at Iris.AI—one of Europe’s hottest artificial intelligence companies—hopes so. She said as much speaking at the CogX AI festival yesterday, part of London Tech Week.

The entrepreneur experienced first-hand the frustrations of opaque academia while at the helm of her previous energy startup Pinexo. “It took us nine months to find the right researcher who was just three blocks down the road,” she tells Forbes.

This is why she started Iris.AI in 2015, a semi-automated scientific review tool that’s today used by everyone from Denmark’s Leo Pharma to the University of Helsinki (this uses artificial intelligence to help scientists tackle “information overload” by pulling together precise reading lists for them).

Now, having announced a €2 million ($2.4 million) fundraise at Finland’s Slush festival in December, Schjøll Brede is building a “blockchain for scientists” to disrupt the traditional publishing model.

“The conglomerates buy up all these journals, bundle them together and sell them back to universities at a ridiculous price,” she says. “It’s a lazy business model, and in a digital world it doesn’t make sense anymore.”

Project Aiur

Schjøll Brede’s new side project is called Project Aiur (a name which references the hit PC game StarCraft). Its aim? To use the blockchain to support a transparent AI peer review and publishing service with its own online economy.

Individuals can contribute in many ways, building useful research tools, publishing their own studies to the platform, peer reviewing work, and training the platforms AI to automate the process.

In return, they earn “Aiur tokens” which can be spent on the use of research tools, including those which allow them to compare their papers with all other scientific research on the system (this today includes over 130 million open access studies, and is expected to grow over time).

“We call this tool the semi-automation of the peer review process,” explains Schjøll Brede. “It helps you spot any constraints you might have missed, any other papers contradicting you. You might still be right, but you can now be aware of it and address it in your paper.”

Schjøll Brede has already soft-launched Aiur’s €10 million ($11.7 million) token sale to 8,000 researchers who’ve worked on its machine learning, with a public token sale due to open later this month—this has a lower cap of €6 million ($7 million) and upper cap of €50 million ($59 million), says the entrepreneur.

No one will be allowed to hold more than 2% of tokens (the Iris AI team will initially hold 50%, but will redistribute this to the community within the next 18 months). And those who want to stick with traditional of paywall publishing can simply pay to access Project Aiur’s features, with their payments increasing the value of its digital tokens.

Tracking “the truth”

As an open source platform, it’s not just scientists who could potentially benefit from Project Aiur, which has been built as an independent non-profit extension of Iris.AI (a bit like the Linux-Red Hat model).

Schjøll Brede says she’s already in “close dialogue” with Unit, an offshoot of Norway’s Ministry of Education, about how it could streamline its due diligence processes before funding new research. Tools could also be built specifically for VCs, patent experts, or even hedge funds, she predicts.

“We need this engine to exist, but we do not believe a company should be the owner of it, because you’re essentially talking about finding the ‘truth’,” says the founder.

“Even if we see ourselves as the good guys taking on a very powerful academic publishing industry right now, we might not be the good guys five years from now.”

Being the “good guys” hasn’t been easy in the billion-dollar journal publishing business of course, and Schjøll Brede says her team’s focus on impact has been the hardest part of her startup journey so far (although having supportive investors like Nordic Impact has helped).

The big dream is to create a world where cutting-edge information becomes more accessible to everyone, she says: whether you’re looking up the latest information about a rare disease, or advice on what you can grow in your garden.

“There is so much in our world of work that is just incredibly time-consuming and painstaking,” says Schjøll Brede. “What we really want to do is take away that excruciatingly boring part so that people can focus on coming up with new solutions.”

Elsevier may have survived one fatal Forbes prediction. Can it possibly survive two?