In February 2017, Google released Perspective, a new API that has sent shock waves through the internet.
Let’s check out what it is, how it will be used, and its potential repercussions for search marketers.
Perspective, Google’s new API, will take any phrase and tell you how toxic it is. By typing a phrase into the tool, the API will quickly analyze your speech and determine how toxic the phrase is. Complex phrases are suddenly reduced to “61 percent likely to be perceived as toxic.”
Now, common sense might tell you that this is a mammoth task, and it truly is. A system with the ability to distinctively alter the whole landscape of the web is now available.
Perspective is built on top of Google’s AI, so it is reasonable to suggest that this system will continue to grow and learn, but it sets a worrying precedent.
Where Will This Be Used?
Do you run a large blog with far too many comments? Then it could well be worth looking at this system.
The Perspective API is a simple way for your CMS to flag any comments that it deems to be too toxic. It would take the sorting out of the process and simplify your workflow while keeping the conversation civil.
Companies can receive hundreds of reviews and far too often these reviews are left to rot. Reviews should never take a back seat, you need to be on top of these things and demonstrate value beyond purchase.
This API could allow you to prioritize. If you are able to see that these posts are more toxic than 90 percent of your reviews, get on those and get them sorted, deal with thanking the others for their purchase later on.
In fact, the New York Times is currently building an open source moderation tool making use of this API. Interestingly, I quickly added a few parts of one of their articles on Donald Trump into Perspective and was presented with a toxicity score of 21 percent, which is a high score for such a large portion of text.
Could we be on the precipice of a world in which journalists have to scale back the truths in their work to satisfy Google?
What’s the Potential Problem with Perspective?
Your freedom of speech is too important to be left to the whim of a simplistic system like this.
The API in the wrong hands could lead to all sorts of issues.
Could your posts on Facebook be hidden if they’re deemed too toxic?
Did you know that the name used in the term “I hate xxx” makes a difference to the perceived toxicity?
Let’s try it with a few names:
Now, Let’s Go a Few Steps Further
What about if this were included in AdWords, that not only do you need to ensure that your text contains the correct number of characters, but is also non-toxic?
AdWords specialists, like myself, would really struggle in some industries to find that ideal phrase that works and doesn’t offend but also converts. That’s a lot to bear in mind when trying to create the perfect PPC ad.
What if the API was included in YouTube or into Google’s search algorithm?
YouTube now auto-generates its subtitles by deciphering the speech within videos. Couple that technology with the ability to create a perceived toxicity value and you suddenly have a situation in which an advertiser is able to decide a “maximum toxicity” that they would allow their ads to be shown alongside. Additionally, this could be used as a negative ranking factor on YouTube for videos which are perceived to be toxic.
If the Perspective API is included into Google’s search algorithm, we’d find ourselves in a future in which perceived toxicity is a ranking factor for certain phrases. We begin to self-censor and the web (and by extension, the whole world) becomes a more vanilla place to live.
We, as SEO specialists, begin to tone down any text, running it through the perspective API in the hope that we can hit an “ideal” level of toxicity.
I won’t go into the morality of it too deeply, but I’d be remiss to not at least acknowledge it.
Yes, this has the potential to make the web a “safer” place. That said, it also has the ability to censor en masse.
As someone who grew up in the “Wild West” days of the internet, lawless and carefree, this tool worries me. As a tool that was built to protect free speech (Google’s words), I’d suggest that this could be the beginning of the polar opposite.