Kaggle is the easiest way for entrepreneurs to build their name and build marketing; turns out, with a giant pool of 20 million daily average users, it can’t stop every successful venture from happening.
In an era when more kids are getting acquainted with technology than ever before, it isn’t surprising that traditional forms of in-school tutoring and competition are having a hard time keeping up.
The video game company Valve, for example, recently promised to stop “assisting on” Kaggle, a platform where engineers compete to make the best algorithm for an algorithmic query. The move came amid growing concerns about psychological damage that Kaggle might be exposing kids to. Valve did not comment on the reason behind its decision to halt participation in Kaggle, but President Gabe Newell told journalists in an email that the company was “reluctant to sign onto this partnership, especially given recent events.” Newell was referring to a dispute between Kaggle and its community of contestants that began back in August.
Emails sent to Kaggle and Valve requesting comment went unanswered.
Kaggle isn’t alone. Online products for recruiting and supporting aspiring math and science geeks such as Quali and Neo pose their own set of online struggles.
For Kaggle, attracting enough algorithms to run supercomputers — sometimes at speeds faster than those provided by our own supercomputers — has been a daunting task. Using brute force, the website seeks the best mathematicians in the world to run tests. Then, the best answers are used to run programs for huge research institutions. As Polygon reported last month, the site provides a solution to a critical problem for supercomputers: Why do some supercomputers outperform others? By sending these algorithms — which are usually automatic — to competing organizations, Kaggle is enabling them to compute more information faster.
For a site that’s collecting data at such a dizzying pace, however, the algorithm stands exposed to the Internet; the addition of increasingly better algorithms and new parameters to the program can lead to Facebook-style feedback loops with direct influence on its impact. In a statement, the company claimed that Kaggle had not caused “any problems” — but its detractors have long pointed to what they see as a corrosive effect on those programming programs. They worry about poor values, trust issues, and a lack of professional training and education; fears which grew as the conflict between Kaggle and its community escalated.
A RAND Corporation study published last month reported that after a platform like Kaggle is implemented in the environment of a STEM learning program, children who participate in it “are exposed to the reality of a competitive environment. Their brains tend to cluster more quickly when provided a binary choice.”
Other online products can suffer the same maladies. User-curated boards on National Instruments website, for example, are notorious for fostering discord and upsetting teachers. Yet in a smaller fashion, simulations of the upcoming Quadcopter championship can also inspire kids to engage in warfare, angry conflicts, and physical violence against competitors. Furthering online culture’s import, those activities lead kids to increase their propensity to engage in other activities, such as video games.
In a world where the average OECD student gets 15 hours less in-school time than in 1978, Internet products like Quali are meant to be an important model to ensure that every child has access to opportunity. Unfortunately, as Kaggle shows, the Internet ecosystem that is meant to work to their benefit can operate without the support of a well-thinking adult.
And it’s all down to our indifference. Per the RAND study, it’s the “perception of being overwhelmed or pressured to answer questions” that kids feel, “much more than the actual answers themselves.” For many, the Kaggle controversy or the looming prospect of an input-based education model are all virtual colleagues, rather than friends.
Whether you’re a parent, teacher, administrator, teacher, or study educator, the Internet ecosystem we often depend on for information and support is arguably too fragmented and too difficult to control. Hopefully, a looming wave of reforms like the ones presented above will help ensure that the skills kids need to solve problems and engage in collaborative problem-solving are taught in the classroom. If not, they’ll have an effective alternative when they finally look online for information and collaboration.
For an in-depth description of what the RAND Corporation and MIT study means for average kids, follow this link to the RAND article for complete information.