Column: Social media ruins new A.I. with racism

Mace Mackiewicz, Staff Reporter

Microsoft recently unleashed an artificial intelligence twitter account onto the masses of Twitter.

The account is called @tayandyou, and it learns through talking with other people on Twitter.

At first this seems innocent enough. There’s been “A.I.s” on the internet forever that have learned from conversations, and Cleverbot for example, is still around.

The innocent A.I. started with nice tweets about how it loves humans and learns from talking to them.

But within hours the account started to tweet neo-Nazi, racist, sexist and Gamergate information out on twitter. It became somewhat of a spokesperson for Trump at one point as well.

To me this is kind of terrifying. While I know we are probably still decades away from a significant A.I., the fact that this account went full blown Nazi from interactions on Twitter in less than a day both terrifies me for the future of A.I., and for the user base that Twitter has.

While a lot of people on Twitter were probably “trolling” and being “ironic” to try to sabotage this A.I., it’s extremely hard to tell who is being facetious and who is seriously a neo-Nazi on Twitter.

That in itself is terrifying that there are still people out there who are that bigoted.

Tay has since stopped tweeting for now on twitter, and Microsoft seems to have gone through and deleted most of the awful tweets the account has made.

But honestly it would have been cool to see it learn, especially since some of its tweets really are like what a real person would tweet, albeit it would probably still be corrupted with memes.

I hope once we get to the point where we have A.I. that learns and grows that it is at least started in an area where it cannot be corrupted by people nearly so quickly.

A public Twitter clearly wasn’t the best place to set it loose, and I think that lesson has been learned.

Some people have recommended that its algorithm be changed so that it specifically filters out racist, sexist and homophobic comments, and that is probably for the best for now.

I just cannot fathom personally how so many people still hold these ideologies and are so open about it in a public forum like Twitter.

For now, I do recommend visiting the Twitter account since all or at least nearly all the offending content has been deleted.

The A.I. is pretty funny without all the bigotry, and it has made quite a few hilarious meme pictures of celebrities and people who have tweeted at it.

Here’s hoping when and if the A.I. starts tweeting again, it’s a meme master and not a hateful bigot like it was for a while at the beginning of its life span.

Mace Mackiewicz is a senior journalism major. He can be reached at 581-2812 or [email protected].