Please Confirm You Are Not a Robot


Warning: Undefined array key 0 in /home/digitald/public_html/courses/wp-content/plugins/radgeek-FWP---Add-Attribution-70acf52/add-attribution-feedwordpress.php on line 363

Thanks, I had to make sure that no bots would read my blog, you know? CAPTCHA is something we most definitely have all seen online. A regular CAPTCHA usually requires the individual behind the screen to type in the displayed text from a distorted image. Despite seeing it, a common internet user most likely does not have a clue what it is used for. First of all, I should have asked what is it being used against? Robots…duh… Well, let’s take a closer look, because every time I am asked to fill out a CAPTCHA, for some reason, I think of an actual robot somewhere sitting behind a screen, trying to complete the required task.

To those who think bots are a modern concept, keep in mind that the first CAPTCHA, according to Captcha.net, was invented in 1997. This makes sense, since even during the early time of the internet, people were trying to bend the rules to their own benefit. A few commercial uses followed in 2001, with PayPal contributing significantly to CAPTCHA’s fame. CAPTCHA was used to prevent a variety of actions: financial fraud activity, creation of free emails for the sake of sending out spam, submitting comments for the sake of influencing the political opinion of the people, and many others.

But while the concept of CAPTCHA does sound very useful, according to the Huffington Post article, many have complained about it. People with certain level of vision disability were unable to follow through the CAPTCHA requests, and were filtered out along with the bots. Furthermore, since CAPTCHAs are computer generated, some can come across as offensive or just simply ridiculous requests. So while CAPTCHA is crucial to today’s internet use, it is important to constantly work on improving it.

 

Posted from DIG 101 Blog by Andriy M.

Strunkbot and Whitebot


Warning: Undefined array key 0 in /home/digitald/public_html/courses/wp-content/plugins/radgeek-FWP---Add-Attribution-70acf52/add-attribution-feedwordpress.php on line 363

Samuel Woolley’s “How to Think About Bots” provides fodder for important ethical conversations around culpability and digital autonomy, but also for philosophical conversations around authorship and art. I was vaguely aware that bots routinely produce articles and writing, but I always thought it would be easy to distinguish such articles as “bot writing.” But when I took the New York Times quiz that Woolley cites, “Did a Computer or a Human Write This?”, I earned a devastating 3/8 (apparently better than 8% of players). All this is to say that bots are incredibly successful at reproducing patterns of human writing and discourse. It really makes me wonder why I have spent my college education programming my brain with the rules of Strunk and White, agonizing over word choice and split infinitives, when a bot can produce similar-quality writing at a much faster pace.

Indeed, even the kind of carefully crafted speech called for in response to a tragedy may be predicted and replicated by a bot. My friend and fellow Davidson student Arianna Montero-Colbert created a bot that generates institutional responses to tragedies—you know, the kind of campus-wide email we get from CQ after a devastating shooting or police killing that makes vague calls for “unity” and “dignity” rather than any kind of coherent political action. Arianna’s bot is scary good at imitating those detached, performative responses, and the results are chilling . We might consider tragedies as occasions that call for some novel or groundbreaking commentary, but Tragedy Bot shows us that even CQ’s eloquent prose adheres to discursive patterns, tropes, and clichés that can be replicated by a bot.

from Arianna Montero-Colbert’s “Institutional Response Generator”

 

Posted from Digital Studies by Hannah L.

The Future of AI is OpenAI


Warning: Undefined array key 0 in /home/digitald/public_html/courses/wp-content/plugins/radgeek-FWP---Add-Attribution-70acf52/add-attribution-feedwordpress.php on line 363

As we look into bots in class, the first company that comes to mind in regards to artificial intelligence is OpenAI. Backed by Elon Musk, and many other funders like Amazon, OpenAI is a Machine Learning research company whose goal is to catapult our current intelligent technology far beyond its current level. What’s crazy is that the company actually accomplishes this by letting the computers teach everything to themselves.

Some of OpenAI’s most successful projects come from their self-play learning system: where apart from telling the computers when they are doing a good job, the computers themselves are left to figure out how to accomplish a task.

One example of this is OpenAI’s Dota 2 bot. Dota 2 is a Multiplayer Online Battle Arena game that has a huge competitive following. The bot itself is unlike the usual bots in the game because it isn’t given any sets of moves beforehand. The bot has to test the water for itself. It learns it has inputs, then that each input has an action, then that there are enemies that do harm, and positive things to do to combat attacks, etc. Until, like what happened around August 16th, 2017, the bot is perfectly capable of playing the game on it’s own.

It is still learning, though, and through more trial and error can only get better. You can read more about the specifics of their task here, but motivating the bots to use tactics that professionals use has given them an edge where they can beat even some of the best players in the world. One thing that is unique that is explained in this video is that the bot even learns from what the opponent is doing in real time. And, because in development the opponent is always a copy of the bot, it always has an equal competitor to learn from. Using pattern recognition and pre-developed strategies from previous generations of the bot, it develops ways to climb its own ladder of success.

Machine Learning is an incredible technology which, thanks to OpenAI and other frontrunners in the field, is helping us understand and progress society today. Although scary, this technology can be used for even more practical applications like prediction of spread of diseases, halting the spread of viruses, and more. It’s even starting to be put into our cars to make commuting safer and easier. Applications for Artificial Intelligence and Machine Learning are everywhere, and are setting humanity up for using the technology to make slight improvements to our lives in the future.

for those curious, here’s actual gameplay:

Posted from DIG101 by Tucker C.

Are Bots Becoming Human?


Warning: Undefined array key 0 in /home/digitald/public_html/courses/wp-content/plugins/radgeek-FWP---Add-Attribution-70acf52/add-attribution-feedwordpress.php on line 363

When reading How To Think About Bots, I realized that most of the things that these bots were doing were considered to be human actions. They were buying things of the dark web and acting independently of the humans that made them. That begs the questions can bots be considered independent of their creators? Or maybe even human?

I think these are important questions to ask, especially in the age of rapid progress in AI and bot technology. It might also to be asking certain questions as to the authorship or responsibility of works of literature or actions in a certain situations in which the bot generated literature or bought drugs on the Internet. This idea of authorship relates to my DIG 220 class in which we have pondered whether or not bots should be credited as authors or the people making the bots should.

It is interesting though to explore the responsibility of the bots for buying drugs. For example, in How To Think About Bots says that even though the drugs that were bought were in ownership of the people who made the bots, “our lawyer and the Swiss Constitution say art in the public interest is allowed to be free” (Power). But isn’t this a loophole in the system? Honestly, I don’t think that this should be able to happen legally. It is a sly way to buy the 10 ecstasy pills, which are illegal, but since it was a bot, it is not illegal.

In think that these scenarios illustrate the way in which technology are becoming more and more like human beings in the way that our actions are becoming more and more similar as time passes. My question to you is: should they start to be treated as people?

I think this image illustrates my point of how I think that bots are becoming more like humans and should be treated like humans because of the way technology is advancing

 

Posted from My blog by Will M.

Smarterchild and Eldercare Bots


Warning: Undefined array key 0 in /home/digitald/public_html/courses/wp-content/plugins/radgeek-FWP---Add-Attribution-70acf52/add-attribution-feedwordpress.php on line 363

I’ll never forget my first online account. I was in fourth grade when my friends told me about AIM instant messaging. Having grown up in a house with no other kids my age in walking distance, opportunities to talk to my friends outside of school during the week were highly limited if those friends were not on my baseball team. Naturally, I loved AIM and spent hours online chatting with my friends. However, there were times when I would get online to find that none of my friends were on at that time, and that’s when I discovered Smarterchild.

A sample conversation with Smarterchild (courtesy of Whiskey Riff).

Conversations with Smarterchild were interesting (especially to a bored 10-year-old) because of how human it sounded some times, and how clearly artificial it sounded most times. Either way, it usually made me laugh for a bit until I got bored and signed off. It entertained me, but never for longer than 10-15 minutes.

In his book The Industries of the Future, Alec Ross explains how the need for more eldercare in an aging population is driving Asian robotics companies to develop robots that can take care of grandma and grandpa. While such firms have successfully created robots that can perform a multitude of tasks for their dependents, developers are still struggling to create robots that can connect on an emotional level. However, as the AI that powers these robots advances, they will learn more quickly than ever, and will soon be able to mimic human emotions.

Courtesy of Dale Thomas on Quora.

Posted from DIG101 Blog by Noah R.

Can bots be trolls?


Warning: Undefined array key 0 in /home/digitald/public_html/courses/wp-content/plugins/radgeek-FWP---Add-Attribution-70acf52/add-attribution-feedwordpress.php on line 363

Last week in class, we discussed internet trolling at length. Essentially, these trolls are people who engage negatively with others online just because they want to. I found similarities in how Samuel Woolley talks about bots to how we talked about trolls in class.

The first thing that made me think of this was the story of a bot sending a death threat on twitter. A twitter user made a bot that took chunks of his tweets and rearranged them into new tweets. One of these new tweets sounded a lot like a death threat, and the Dutch police came knocking. Many examples of trolling on the more malicious end of the spectrum include making death threats to a target, like in the Gamergate “shitshow” where women received violent messages from online trolls.

The article’s discussion of bots also reminded me of trolls because of the difficulty in moderation. For both, the issue of free speech comes into play, because to limit bots or trolls could also limit free speech as a consequence. As Woolley says: “rumination on bots should also work to avoid policies or perspectives that simply blacklist all bots.”

The problem with comparing bots to trolls is that bots lack the sentience that trolls possess. One of the key points of a troll is that they get a kick out of what they’re doing to other people, and that is their motivation for continuing to troll. Bots, on the other hand, don’t have feelings, and therefore can’t get satisfaction out of what they’re doing.

via Twitter

Can a bot be an evil troll if they’re not even aware of what they’re doing? Is sentience a critical part of being a bully? These are questions that further investigation, persecution, and revision of bots will hopefully answer.

Posted from Digital Studies 101 by Alivia

Let bots be bots and humans be humans


Warning: Undefined array key 0 in /home/digitald/public_html/courses/wp-content/plugins/radgeek-FWP---Add-Attribution-70acf52/add-attribution-feedwordpress.php on line 363

Upon searching Twitter for popular bots, I found an interesting one with the handle: @DearAssistant. Ask the bot any question, and the algorithm that drives the bot will respond to you with an answer.

But wait. Don’t we already have something for that? Last time I checked, it was easier to get information with a quick google search than it was to use a Twitter bot for answers. Heck, I could just ask Siri some of these questions, and she should be able to get back to me within a few seconds.

Yet, this was a popular account (at least until it stopped tweeting in 2016), and the bot accumulated more than 4,700 followers. Clearly, the simple interaction with bots like these is what drives their popularity and existence. It’s not always the bot’s usefulness; I’m sure the user in the above example knew what day it was but still wanted to see the bot’s answer.

In this sense, How to Think About Bots hits it on the head with the idea of “botness.” The way a bot is “not convincingly human” is amusing to us.  If a human was asked “current time in London?”, I highly doubt he or she would respond down to the second like the bot did. That preciseness ( what makes a bot, well, a bot) is what drives us to adore these creations.

Look at this same user’s response to another question.

I remember a service called “Cha Cha” that you used to be able to text for answers to any question you may have. It was completely free for awhile, so long as you didn’t reply to any of “Cha Cha’s” promotional messages.

This service differed from today’s Twitter bots like “Dear Assistant” in that humans were the ones replying to our messages. Is it any fun to ask a human the time of day? Not really. But is it fun to ask some random human “Where’s Waldo?” or “Who is Grant Labedz?” Definitely.

That’s what the service became for me and my friends. It was intriguing to see what kind of witty, intellectual, creative responses these humans could come up with on the other end of the line. Some would actually have fun with it too and feed us with surprising responses that made us laugh.

In a sense, we expect different things from humans and bots. Bots are exciting because they’re bots. They fulfill their functions in a “bot-like” manner, and to us, that’s cool enough. From humans, we expect more. We want creativity, wit, and sometimes more than just information when asking a question.

Posted from Digital Studies 101 Blog by Grant L.

Trolls vs Bullies


Warning: Undefined array key 0 in /home/digitald/public_html/courses/wp-content/plugins/radgeek-FWP---Add-Attribution-70acf52/add-attribution-feedwordpress.php on line 363

Trolls control the internet and wear many different masks. There are some that aim to personally attack and insult others, and there are some that look to annoy and or mildly inconvenience. I personally have been lucky enough to have never faced a “troll” whose intent was to offend me. Granted I don’t have a big persona on the internet, which makes me a pretty irrelevant target for trolls. In my experience trolls have always been a part of my experience when playing games online. I am not a huge gamer, but I dabble in the occasional game here and there. The first time I was ever trolled was when I snuck onto my brother’s computer and played the multiplayer of Halo: Combat Evolved. 2 guys were standing at the spawn just shooting everyone on their own team when they respawned. They would call you a name in the chat once you died and were being just generally annoying. It was enough to discourage me from playing games until I was older and understood what they were doing.

This to me in hindsight was frustrating but harmless, as I think most “trolls” are. Usually they are in it for a laugh, though it comes often at the price of another person’s inconvenience or mild frustration I seriously doubt most people have true malicious intent. I think that when someone has the goal of mentally hurting and or attacking someone personally that there should be another word for their actions. It is no longer the mildness of a “troll” but rather the aggression of a bully. I feel like calling the actions of those who delivered threats and attacks during “Gamergate” the ones of “trolls” really makes the whole situation too lighthearted. When I hear trolls I think of those guys who ruined my Halo game, not of misogyny and brutality. I agree with the article in that we need to reevaluate how we act online, but I think the first step comes with recognizing the impact that “trolls” and or “bullies” have on people’s lives outside the digital world.

http://ei.marketwatch.com//Multimedia/2016/07/14/Photos/NS/MW-ER478_trolli_20160714084402_NS.jpg?uuid=a4a30a84-49c0-11e6-82a4-0015c588dfa6

 

Posted from My blog by Patrick H.

Lab 3 – Animated GIF


Warning: Undefined array key 0 in /home/digitald/public_html/courses/wp-content/plugins/radgeek-FWP---Add-Attribution-70acf52/add-attribution-feedwordpress.php on line 363

 

Reaction gif

These are some of the most commonly used gifs in today’s worlds. Finding a counter-gif for this wasn’t too hard due to its regularity. I decided to take a different path with this though as the gif I chose as my counter is a reaction gif itself. Ironic I know, but reaction-gifs are all about context and this gif is making fun of that context. To elaborate, the internet loves trends and gifs are no exception to this. Currently Reddit is packed full of Michael Scott gifs reacting to a caption about how people’s wives are already starting to take out Christmas decorations. It is a relatable idea for many and his simple and or over-the-top reactions seem to fit it perfectly. Someone had the great idea of posting this gif of pam (image 2) with the title, MRW (my reaction when) I have no wife or girlfriend to take out the Christmas decorations early… or Christmas decorations to even take out.

Cinemagraph Gif:

These are my personal favorite out of any gif category. They are visually captivating and practically utterly useless. There is often not a very humorous meaning to the gif. It is meant as a piece of art. That is why I think Will Farrell’s gif couldn’t be more fitting. It tears down the fancy aspect of a Cinemograph. We see in the first gif a picturesque dinner table with the endlessly swirling glass of wine. In the second gif Will Farrell is in a massage chair, crying, and failing to get the wine from his swirling glass to his mouth, hilarious. In my mind that is what a “gif” should be. It should tell a story or describe a moment. The cinemagraph creates a moment, its purpose isn’t to capture a scene we can all relate to. Though I do thoroughly enjoy scrolling through these cinemagraphs I feel like they are the snobs of the gif world.

Wobble Gif:

The wobble gif can often be a parody of other gifs due to the fact that it can combine multiple gifs and or scenes into one. I picked the inception gif because it is such an iconic and frustrating scene that many people have often done some funny parodies of it. The gif loops endlessly with the top always wobbling every half second. People like to put captions such as, “wait for it.” Why you ask? Because they find it hilarious to know that you will stare at this endlessly looping gif with the expectation that something will eventually change. Even that wasn’t really a parody because the gif itself was unchanged. The second gif shows Richard Branson in a high speed car chase using some future device to help eliminate the cars in pursuit. Some creative animator realized that his device looked similar to a Gameboy so they edited in a Gameboy in place of the actual piece of technology. So now he is simply playing an intense game of Mario-Cart during what is supposed to be a suspenseful chase scene. I love this gif because it is able to totally transform the clip from serious to comical with one simple edit.

Sports Gif:

These are famous for two reasons, highlights and fails. That’s about all you get with a sports gif typically. Sometimes there is an exceptional fan reaction, but generally speaking they are, to no surprise, sports oriented. The first gif I chose is a classic play where the cowboys fail to touch the receiver down and give up a touchdown against the giants. The second gif is of Homer Simpson fading away into the bushes behind him with a cub’s shirt on and reemerging with an Indians shirt on. It is meant as a way to show the loss of faith in the losing team and the shift to the winning on. It is a funny comment and a good break from the standard sports gif. I think that as sports gifs go this is rather counter to the norm as it is breaking allegiances which is a big no go when it comes to sports.

Fandom Gif:

The nerdy brother of the gif world is the Fandom gif. I picked Game of Thrones as my topic because of its prominence in today’s culture. Even with its popularity it can be hard to use a lot of the show in gif form as people have a hard time separating a moment from the show as a whole. When I see a clip from game of thrones I remember everything else that was going on in the show and I cannot really separate it to apply it to a situation in my own life. Without that quality the gif either has to present a joke and or remind you of a classic scene. The first gif is an iconic scene for the show of the Night King, doing what the Night King does. The second is a totally different scene with a very different meaning. The gif of Tyrion is meant to make you laugh and it can be used as a reaction gif or just for a laugh. This to me makes it counter to the Fandom gif because it serves no purpose in really highlighting a moment in the show. It can be removed from all context and still manage to get a laugh or at the very least feel relatable in some way.

 

 

 

Posted from My blog by Patrick H.

Origins of Trollface?


Warning: Undefined array key 0 in /home/digitald/public_html/courses/wp-content/plugins/radgeek-FWP---Add-Attribution-70acf52/add-attribution-feedwordpress.php on line 363

Personally, I have never really understood trolling. It is something I did try to oppose at first, but since that approach simply adds to the mess, I learned how to ignore it. Reading both “We’re the Reason We Can’t Have Nice Things on the Internet” and “Privilege Makes Them Do It — What a Study of Internet Trolls Reveals” forced me to research if there are any hidden benefits online trolls receive for what they do.

 

Even though I was not lucky enough to find the overall definite motives of trolling, I did discover something very interesting.

Trollface

We have all probably seen this meme. If not, meet the Trollface. For me, and probably for many people around the world, this caricature of a smiling face has been the logo for anything troll-related. However, not many know of this meme’s origin. According to one of the articles on the Know Your Meme platform, Trollface was introduced to the world on September 18th, 2008. This was very surprising to learn about, since the concept of trolling and troll-like behavior, tracks back a long time before that date.

Another very interesting fact this article pointed out was about  the author of the Trollface. Throughout a few years the man has made over $100,000 from his creation. Which, at the time of the number recorded, was approximately $15,000 a month. This was especially surprising, since despite the anonymity and the violation of many rights and morals, the meme ownership rights have been widely acknowledged, and, at least partially, paid for. Turns out trolling related things can grow into a business, and in fact, bring quite a bit of money to the people behind them.

 

Posted from DIG 101 Blog by Andriy M.