1 April

I Was Offered An AI Software Experiment To Block Hatred, Chilling And Shockingly Effective. I Can Have It For Two Weeks. I Won’t Be Using It

by Jon Katz

Long Ago and Far Away, I used to write for Wired Magazine. One of the people I worked with has become a well-known software engineer in Silicon Valley. We were both relatively young and realistic when the Internet blossomed, and we were convinced it would be a blessing for freedom and democracy. Finally, information would be free.

To some extent, that dream is true; to another, it often appears social media is chewing up democracy and threatening it severely. Online hatred has become a political and personal weapon, increasingly used to punish disagreement and free thought. My friend and others are now claiming AI software can turn that tide.

Being young and idealistic, we didn’t imagine the greed and indifference of corporations, who allowed their websites to be cesspools of hatred, rage, and bigotry to draw an audience.  We didn’t foresee that this anger and rage would grow and threaten our democracy rather than nourish it. Writing in the open on the Internet is now too often asking about social bullfighting.

I’m skeptical of my friend’s idea.

Reading my blog, he’s followed the hatred and cruelty that seem to be blossoming everywhere and on my blog at times, making me a fascinating proposal. He wants me to use new and experimental AI software designed to stop and block hate messages. It works simply, he says. The AI software is fed thousands—even millions—of hate messages and the e-mail, fake or real names of the people who send them.

They do this by collecting and storing the names of people who send cruel, threatening, or vicious messages. They also collect what the messages look like, feel like, and sound like. When the software is fed and activated, it automatically and instantly identifies and trashes or destroys these messages instantly.

I would no longer see them, although they can be stored if I ever wish to read them. He wants me to try this software for up to two weeks and report my experience.

I thought about this for a few hours, but no longer than that.

I thanked him for thinking of me and declined his offer.

As you know, I believe in confronting and sometimes exposing hate mail. I wouldn’t say I like getting it (who does?) but thinking about it, I decided that what the AI software should do is precisely what I am beginning to do and should do myself.

Instead of using software, I would block or destroy these messages on my own and continue with my work rather than surrender them to the software, which is so intrusive and technical that no one, including me, would know how it works, how it gathers all this material, or what it would ultimately do with it.

My arguments with haters (yes, you, Jullie) are well known and often tiring and frustrating to me and the innocent bystanders who must listen in or hear wanton cruelty and hatred.

But I am getting there, and no software will strengthen and empower me more than me. I’m not turning this task over to AI software; I will never be sure I have the strength, skill, or confidence to handle this issue myself if I do that.

Hateful messaging has not prevented me from doing my work, writing what I want, or loving my blog, my life, and the many excellent and non-hating people who read it.

Yes, I know this is how the Nazis started it, but I am not prepared to equate what is happening in America to the Nazis and what happened in Germany. I think that’s going too far for me.

I installed much more moderate software that allows me to delete hateful messages instantly and, with one button, ensure that no one sending such a message will ever get posted again and move their messages straight to the trash.

It’s straightforward. The software identifies names, languages, e-mails, and sources from which they are sent. It does not block, delete, or eliminate people who disagree with me, dislike me, or wish to challenge me on my thoughts. I can delete the haters if I wish, but I don’t. Those are the messages I want: thoughtful, intelligent, and civilized. Those are the messages I am getting now.

This is my work—writing in freedom, exchanging ideas, hopefully getting people to think, and hoping they will return the favor.

My software has blocked hatred 100 percent since I installed it, and I glance at it once or twice a week to ensure innocent and good-meaning people are not being blocked by mistake.

I don’t seek Nirvana. I love having some tension on the blog; it suits my thinking. I hate having hatred on my blog; it obsesses and detracts from me and my ability to think. And I don’t want an invisible screen between me and those who read me, no matter how much some might hate my guts.

There is an Orwellian element to this.

The haters will only know their messages never appear in my blog comments or appear in my e-mail. They do not know; they are banned, blocked, or banished unless I tell them or unless they tire of yelling into the sunset without the satisfaction of knowing they hurt or dig some damage. In my experience, they feel the best solution is quitting and going away. Haters need an audience to hate; silence does for them what sunshine did for Dracula.

But it has worked for me. I don’t miss the hate; I don’t miss the hatred it pulls out of me; I am more accessible than ever to write what I want and share it with people who want to read. Slowly, day by day, my dream is coming true, and if anybody gets credit for it, I do, and my readers do too.

As tempting as it is, I don’t want AI software to do this for me. I want to prove to myself, and yes, the world, that this can be done by ourselves. Hatred flourishes only when it hurts and is listened to. I can’t guarantee this will be a kinder and more compassionate world.

But it’s a good start, better than letting this hatred and cruelty flourish without challenge or the power of decency and honesty. Good luck, Jullie; I can’t honestly say I wish you well, but if you are reading this, know that you are gone and will never be admitted back.

My message to me is good for you, Jon; you are dealing with this in a mature, realistic, and hopefully compelling way. I told my friend I honestly believed I could do it as well as his software could.

Maybe I’m just arrogant; perhaps I’m a visionary. I won’t be a coward.

We’ll see.


  1. The trouble with your idea is that you alone get to define the difference between “hatred” and thoughtful disagreement. I don’t particularly trust your barometer for that.

    1. Charlotte, thanks for the message, but I’m confused. This is important to me.

      Since I rejected the idea, I’m not sure what my ideas about disagreement and hatred have to do with your comments. I could be wrong either way. Seems like an irrelevant way to do some sniping. If you are trying to criticize me thoughtfully, I’d prefer you come out and say what’s on your mind. Right or wrong, the decision has to be mine, or it can’t work for anyone. I haven’t killed anyone yet. My wife really loves me, and lots of people like the blog. Your message is not hateful in any way, so I’m not sure what you are suggesting. Even if I am a total jerk, which is quite possible, I decide what appears on my blog: hate, thoughtful messages, or cheese on the moon. I don’t have a barometer for this; it’s complex. If you have such a barometer, I’d love to hear it rather than just another poke.

      Which idea is a bad idea, do you think? Turning down AI software. Is that the idea you would trust?

      The blog is mine, and I feel very strongly about it, as is evident. I started it, paid for it, and did the publishing work and photography, not you or anyone else. I decide what is hateful to me or thoughtful to me. I don’t really need your approval and have survived without it in the 17 years the blog has existed. I’m sorry, but you don’t get to decide what feels right to me. I do not feel obliged to post hurtful or hostile messages, whether I am correct or not. Your right is to read the blog or not. That is the most powerful position of all. You are welcome here; I see no hatred in your message. Thanks for contributing in a civil way.

  2. Well done. Well-thought-out process. I admire the conclusion you arrived at. I like the idea of humans being in charge of decisions and choices rather than AI.

  3. * no software will strengthen or empower me more than ME*……….well said, Jon. AI………I am educating myself about it….husband is starting to use it…….. I have nothing against it……but…..the human brain of ours is still the best when used to its full potential., IMO
    Susan M

  4. Interesting that someone would think that you shouldn’t get to decide, on your own blog, what is hatred and what isn’t. That baffles me. This quote by Brene Brown comes to mind “… If you are not in the arena getting your ass kicked on occasion, I am not interested in or open to your feedback.”

  5. I agree wholeheartedly. I prefer to decide for myself as well to maintain a modicum of control over my online life.

  6. It really doesn’t seem like more monitoring and filtering of messages is the answer. Not that I have an answer. It’s clearly true, though, that this is not an open discussion board; it’s a website belonging to one person who pays for it to exist. In that regard, it’s private property. You can’t march up to someone’s house and splash paint on it just because you prefer purple to blue (or for any other reason). And although there are laws against that kind of behavior, I don’t think it’s the laws that keep it from happening. I think it’s that people share a generalized understanding of what’s acceptable and what’s not. This Internet stuff is still pretty new; maybe it just takes a while for people to realize basic decency applies here too.

    1. Thanks, Pete; I appreciate the thoughtful message. The Internet is not new. It’s over 60 years old and has evolved in many ways, good and bad. One thing that has grown is hostility and cruelty, which seems out of control. Monitoring messages is the only answer; technology and the ease of messaging are not controllable without the people who run the sites taking responsibility for the civility of their blogs and websites. Believe it or not, this is beginning to happen, but only on a very small level. Reddit is an excellent example of a sewer site of hostility and cruelty being reworked and asking its members to take responsibility for the messages and topics they use. X, once Twitter is an example of where it is going.

      It is working there. Reddit is one of the best-revamped websites on the Internet, a place for thoughtful news and discussion; it is growing and making money and has eliminated hate groups and hateful people. For the most part, it can be done. However, the big sites run by corporations don’t care about taking the trouble, as Musk and Facebook have demonstrated. There is no generalized understanding of what’s acceptable and what’s not – this is what I am adding to my site: clear boundaries. Cruel and hostile messages will no longer be posted here, and I am responsible for that. A polarized and aspiritual country has no universal or generalized agreement about what hostility is. The humans in charge have to define this for themselves. The Internet has had a lifetime to tackle this problem; it has worsened every year and is at its worst now. I’m afraid I disagree with your feeling that time will eliminate this problem; I’ve grown up with it and see it only worsen. I wish it were that simple. Thank you for your civility and thoughtfulness; you are always welcome here, just the kind of disagreement that is valuable and worthwhile.

      No culture thrives without boundaries, hatred thrives without them.

      1. Wow, Jon. You clearly know nothing about Reddit. It certainly hasn’t eliminated hate speech. One of the fundamental principles of Reddit is that anyone can start a subreddit on any topic. Very, very few posts are ever removed. Users upvote posts they like and downvote posts that they don’t. The theory is that, rather than moderating and removing posts that some single forum moderator deems “objectionable,” community opinion will take care of it. Posts with the most upvotes rise to the top and the most downvotes sink to the bottom. It’s the complete opposite of what you do on your blog, where you and you alone decide what comments people are allowed to see.

        1. Wow, George, I know little or nothing about Reddit and am happy to keep it that way, nor do I wish to argue about it on my blog. I don’t need Reddit to do what I do or for me to do what they do. My comment came from a New York Times article on how Reddit has changed, written by a tech writer. I’ve read it elsewhere, as well. If you have any complaints, take it there. I’m happy to post it here. If you want to argue about it, it won’t here; best Jon


    2. P.S. Pete, there is no law against hateful messages anywhere in the country if they don’t threaten violence.

  7. Jon

    I would gauge my frustration with hate messages for 2 weeks and score it daily. Then try the software for 2 weeks and score that. If you are less frustrated then philosophical thinking about it is not necessary. You are less frustrated /happy so use the software.


    1. Thanks Frank, a good idea, but I’m pretty clear, I don’t want to test it out, I just don’t want it. I’m confident about my borders plan. thanks for the thought.

  8. Jpn,
    I agree that you should be in control of your blog and you should make the decision as to what should be be posted and what should not.

Leave a Reply

Your email address will not be published. Required fields are marked *

Email SignupFree Email Signup