AI and Research Accessibility

Original article was published by Jasmine Grewal on Artificial Intelligence on Medium


AI and Research Accessibility

In what we’ve begun to call “quarantimes”, many are anxious for the end of lockdown. That will probably only occur once we find a viable vaccine. For now, that’s far away, but most still jump to read an article mentioning a vaccine, but why must the public rely on the news and media rather than the published findings of actual studies and research?

Often we get our news and updates of what happens in the world from news sources like the New York Post or CNN. These major news publications are considered trustworthy, but we don’t often consider that these publications usually don’t offer any sources, and they are fully entitled to withhold or embellish stories as they see fit. In other words, we forget they can make fake news any way they please. While articles may be written using scholarly language and bits of evidence, but they may still manipulate articles to serve the interests of investors, shareholders, or their own.

When faced with propaganda that twists the results of actual scientific studies, an obvious option seems to turn to actual publish research. While these studies are easy to find with a simple google search, the obstacle is readability. 88% of US adults between the ages of 16–65 read at PIAAC level 3 or below. This would be more complex than your average news article and would use more scientific jargon than most are used to. If you’ve ever read a journal, you’d see that these would easily hit PIAAC level 5, and for good reason. They can be used as evidence in copyrights, drug patents, and lawsuits. They must be thorough and they can’t leave anything out of their published results.

But while these make sense in the aforementioned scenarios, they make them wordy, and easily boring to someone who just wants to know what the basic specs of a new vaccine are. If you wish to know the possible side effects and adverse reactions in trials, you don’t want to know the exact specifications of the trial subjects and method. But AI could be used to help bridge this gap between the average reader and unbiased information.

Recently several companies, such as Grammarly, have begun developing AI to edit essays to check for typos, grammar, and readability. While these are currently targeted towards school-age children and young professionals, these have great potential in making unbiased research accessible to the average person. Once these current technologies are developed further, they can be specialized to different fields, such as mathematics and medicine.

Sources: