The Internet has given power to everyone. So how one uses it completely depends on the user. But if you closely watch social media, the internet, blogs, there is so much toxicity in it. People write whatever they feel however they feel without even choosing the appropriate words.
These people do not even realize that the things that they are writing or expressing are being directly hitting someone. It can be a personal comment or a comment on a community or whatever. But these negative comments are known to trigger events or even online wars.
But the earth has a human population of 767.35 crores. And more than half of this population uses social media or the internet in some way. This gives the population the liberty to explain and express themselves in whatever way they want. The words like toxic content, trolls, negative comments, personal attacks need to be constrained. And the internet needs to find ways to control them.
Now, educating the population or defining rules surely has not worked. So this is where AI steps in.
So, How AI Will Help?
AI has come a long way since the time it developed its name. So why not make it more intelligent and define algorithms to be put into use to defy toxic content. AI algorithms can scan and put aside the list of words that are not correctly put to use. This could contain a list of words that if used in the comments or any section of the internet could be sent under the scanner.
Well, there are algorithms defined by Google and some other tech giants and were put to use. They did become successful to some extent and gave the area more exposure to be explored. Engineers and developers are working actively to build tools and programs that will not hamper the right to speak but put some sort of filter so that the comments are stable.
And this is only possible through AI.
Moreover, algorithms have come a long way from binary numbers. From switching on lights and ACs to robots doing surgeries, things have evolved rapidly. So when all of this is possible, toxic content can certainly be controlled by AI. Artificial intelligence can design robots or algorithms that can put aside rude comments or give a warning saying that the content is inappropriate to be put out on the web.
Or before posting a comment it can suggest you some sober words by eliminating the words that will instigate negative sentiments. Or it could also give a toxicity score to the content being used.
Like there are tools that test plagiarism there can be tools to test the toxicity and tone of the writing. Plus if the same IP address, email address, domain name, or whatever is being used repetitively the robot could scan it down and make sure that it does not happen again.
Well, the cybercrime department does that but only after the crime or harsh comment is being reported. The AI could detect and make the repairs well before the damage is set to be done.
Failures till date
Tech giants have been trying to keep up with defining AI technology to be used to define toxic comments over the content but have not taken the giant leap, yet. For example, the sentence “ I write a stupid essay.” is portrayed as a negative comment because of the word “stupid.” Because the word stupid is on the negativity radar.
So not only the words but their usage in sentences have to be defined. The sentence may be negative but it should not invite trouble. Designing such an algorithm is taking a toll on the tech giants.
Basically just defining the group of words is not the solution. Where and how they are being used is also to be understood by the robot to be scanned. Post this, the content could be defined as fanatic or harsh.
Research Till Date
Tools have been developed and identified to let the AI read the comment before it is posted on the thread. But it led to failure as “ I am a black woman.” is not a racist remark. But the robot did identify as a racist remark because of the usage of the words black, woman.
So the research team is welcoming everyone who can deal with this issue and get the best algorithm on board. Google hosts the Kaggle competition every year to get in developers from around the world. These developers come up with ideas to deal with the toxicity of the internet with the help of AI.
Well, some of the research has been successful too. For example, YouTube has an option of disabling the comments sections. As a lot of children and a young crowd has access to YouTube, this was a great step to be taken. Some sites disable certain comments and hide them.
They even post a disclaimer saying, comment hidden, and inappropriate language being used. So this shows that research has reached some of its milestones but accurate results are yet to be achieved.
To conclude,
AI defining toxicity on the internet is not limited to text. Even videos, images, emojis used inappropriately need to be controlled. It sounds impossible, but there has to be some way to get the solution to the problem. We do not want events in Washington Dc or Myanmar to be repeated. We do not want people instigated for the wrong reasons using social media.
And we also do not want harsh measures to be implemented by the government as cutting down the use of the internet or disabling social media.
All we need is a perfect algorithm that decodes all our problems and comes up with satisfying solutions. And we know that day is not far away when we will have a robot deleting and not allowing harsh comments to be posted or at least posting a warning that the content is sentimentally hurting.
Author Bio:
Timothy is a lawyer by profession but does a lot of research on the use of the internet. His cases take him on a tour of how people explore and misuse the internet sometimes. He is also passionate about researching, and updating people about the latest developments in the tech industry and frequently publishes his works at https://mycustomessay.com.
