Good Intention, Dubious Approach
Clearly there’s a problem with toxic comments. But assuming they can be constrained in any significant way is as silly as attempts to constrain fake news. Yes, toxic comments, like fake news, can be corralled by AI for us carbon-based units to examine. But three things then happen —
- This system is gamed better
- The human effort is endless until AI replaces them
- When AI replaces humans it will constitute a model for pre-crime detection ala Minority Report (including anti-government speech)
Now, I may be missing something. But given that these problems have existed in other media forever and, short of dramatically changing human behavior (nah), the intelligent approach is to establish access to comments as a revocable privilege. It is, in effect, what now happens here on Medium.
As for fake news, the time has come to develop databases of superior content resources. No one is obligated to use them but, like Consumer Reports, they provide comfort that there’s minimal risk of being misled, scammed, receiving bad, misleading or erroneous information.
Bottom-line: these problems are here to stay. Band-aids will not suffice and attempting to do so could lead to a cure that create new, potential worse unintended consequences