In this month’s “Let’s Talk: Tech in the Arts,” Alyssa and Grace discuss trending news topics that include bias in artificial intelligence, opinions on chatbots, and the recent uptick in hologram concerts.

TRANSCRIPT
[Intro music]

Alyssa: Hello, and welcome to the second episode of our Let's Talk Tech In The Arts series with the Arts Management and Technology Lab. My name is Alyssa and I am the Podcast Producer.

Grace: And my name is Grace the technology and innovative Content Manager.

A: Each month we bring you through trending stories and discussion fields with topics such as CRM, artificial intelligence, marketing, social media, inclusion, fundraising, and much more. Our goal is to exchange ideas, bring awareness and stay on top of the trend. In this episode, we talked about ethical questions within artificial intelligence, the latest on chatbots, and new questions on concerts featuring hologram artists.

[Short musical interlude]

A: Okay, so, um, we actually just got back from a trip called Network New York. This is a trip that Heinz College sponsors for all of the folks within the different programs such as the public policy majors or in our case, like the arts, the arts managers.

G: And the entertainment, yeah, entertainment industry management.

A: Yeah. Oh my goodness. Do you think that you enjoyed the trip?

G: I did. I did. I, actually I, before joining Heinz and AMT Lab, I lived in New York for several years so it was kind of like a homecoming in a weird way. It sort of felt like I never left but was also newly seeing the city for the first time. [laughter] And so yeah, it was really interesting to be there, in a new mindset and a new time in my life, as well as meeting all of the Heinz alum that we got to meet, which was really fun.

A: That makes sense. I mean, for me, personally, like I've always lived kind of close to the city. It's certainly different than actually being in the city and being in, like, with that mindset.

G: Absolutely. There are a lot of things that I was like, Oh, right, I had forgotten about that! But yeah, it's still the same.

A: Yeah, it was fun. It was fun. Yeah. And not only that, but you got to interview a MAM alum at Capacity Interactive.

G: Yeah, Jess Bergson. We had a lovely chat about Google Tag Manager, which, if it's not already up will be up soon. So look out for that AMT Lab listeners.

A: Alright, cool. Hey, looking forward to getting an episode out.

G: Yeah, I'm looking forward to getting it out for listeners as well. A little bit editing to do. [laughter]

A: Alright, let's get amped with AMT Lab.

We're starting off today with artificial intelligence, or AI, and one topic that's recently been making the headlines, bias within AI. So, for anyone's who's, who's taken a programming class before we know that the question of computers gaining a mind of their own and taking over the world via AI has no actual foundation. Apologies to those of you who love 9000 or the Terminator. Anyway, the reason we know this is because computer programs have to be told exactly what to do, but part of the design of artificial intelligence is the information that programmers plug into it. The current issue, however, is that the information that's plugged in may contain bias.

One example that Rachel Adams of the Conversation features is the male and female voice of Bixby, the Samsung virtual personal assistant. If you ask the only male voice on Bixby to talk dirty, he responded with “I read that erosion is a real dirt problem”, while I asked one of the female voices the same question, the response was, “I don't want to end up on Santa’s naughty list”.

G: That’s so different!

A: Okay, well, I say all this in past tense because I actually tried this with Bixby on my own phone and Bixby no longer responds to this question.

G: Interesting…

A: Yeah. So additionally, if you ask any of the male or female voices a flirtatious question, they may respond with various answers, but otherwise all will eventually end up with the same responses regardless of gender. So, it is indeed possible to touch upon some of the bias and programming. However, we see other stories in the news recently, such as ImageNet, right?

G: Yes

A: They have millions of images in their database with different tags and as users recently discovered, these tags assign labels to people's faces sometimes and this has become problematic among representation, diversity, and fairness. Now as a September 17, at 2019 image/net.org has a link to a statement on their main page with updates on how they're going to combat this, including the removal of thousands of images creating this bias. But you know, I'm doing this level of programming and then re-programming, it is undoubtedly going to be a challenge.

So, what do you think Grace? Is it possible to eliminate all bias within AI programming?

G: So, eliminating all bias between AI programming and the results that are shown, that's going to be really difficult to do and social-, social biases, whether it's conscious or unconscious is takes a lot of time to reduce and get rid of completely. And with, you know, as much as we try and get rid of biases like these, we can constantly reduce them, but getting rid of them entirely, I don't see that happening anytime soon. Um, thinking of Bixby; however, that is a good example of trying to reduce biases, at least from the gender stereotypes and gender binary. You could also consider Snapchat’s issue of filters not working on individuals of darker skin tones. That had to do with individuals who were coding were also testing on themselves and as it happened, most of the individuals, if not all of the individuals within that testing pool, had lighter skin tones. And so being able to acknowledge that is the first step. And I think we're starting to see that, like you mentioned with Image Net.

Now Image Net also has somewhat of a slightly different issue as well. I think part of Image Net’s issue was that it wasn't just visual bias, it was also social bias and how those tags were assigned. And from my understanding the artists’ idea was to try and target and bring up some of these discourses that were uncomfortable and really target the way that we think and how certain phrases and certain labels are, are given to people. But I think they started to realize that within that there was this, already bias that was happening. So, I don't know. We’ll… it's going to take a lot of time, I think.

A: Yeah, absolutely. I 100% agree with all of that. And, you know, I'm thinking back to some of my journalism classes all the way back in high school, where the very first thing that we learned is that there's no such thing as no bias in media. Like news, for instance, like, there's always going to be a certain way to present it and you as the author, like, have the ability to pick and choose or even phrase a certain way, and even one word can change the entire sentence.

G: Right. It’s the human element that we can't get rid of. But I do think that there are steps that you can take one, which we've already mentioned, is acknowledging and identifying the bias that you do have. And then secondly, figuring out how can you reduce that bias based off of who your audience is of your technology or the art that's using that technology? I think that is that's a step in the right direction, at least.

A: Yeah, absolutely. And let's see, for Image Net itself, it seems as if the CEO, like from their statement that they posted on their homepage have taken those first steps already. There's actually an interview that was featured on new statement featuring Trevor Paglan, the CEO who helps to create Image Net. He mentioned at the very end of this article with the quote, “The entire way classification is conceived of in machine learning systems is always going to be discriminatory. And that is a feature of the system and not a bug”. But interestingly, like the article itself, it’s title-, it’s titled that his thoughts regarding AI is currently taking us backwards.

G: Interesting. I wonder if he's referencing the sense that because you have that human element, bias will always exist. And I think to a degree he may be correct. I wouldn't say that that is really the best way to consider it, you know, from a viewpoint. But I think it's interesting, too, that he's talking about AI taking us backwar-, backwards. Maybe that's, I wonder why he thinks that.

A: Let's see, well, the thing about artificial intelligence in particular is that at the end of the day, it's designed by humans. And if it's humans that are adding in the bias or have the ability to take it away. That's probably the most important thing to note. So that this way, whenever we do get around to changing the way that AI behaves, like at the end of the day, it's the human element that we need to focus on.

G: No, something, definitely something to consider whenever AI is in the conversation.

A: Absolutely. So, if listeners want to learn a little bit more about bias in AI, there's actually an article that is posted by contributor Laurin Cornwell on a empty dash lab.org which answers questions on this particular subject.

G: So let's get a little more specific in the AI world by focusing on our favorite thing on lots of different websites – chatbots!

A: Oh my [laughter]

G: Chatbots - they've been used to help answer your questions, guide you through museum spaces, provide customer service, book reservations, perform commands, the list goes on and on and on. In the arts world, chatbots like Dot of the Akron Art Museum can give you and your group a personalized tour of the artwork while sparking conversation topics. Recently in customer service, bots are starting to sound human, but something within the cadence is different and that can be a touch unsettling for some. So, let's kick this off with this question. How do you personally feel about chatbots?

A: My goodness, I mean, chatbots- well if I'm talking to my insurance company on the phone, um, then it'd be probably good to have a real human behind that voiceso that I can officially know what exactly I'm doing. But if I'm just setting up a doctor's appointment, that it's nice to, like, not have that human element present where in case, like, embarrassed, I embarrass myself on the phone. But, you know, let's see. I mean, chatbots still certainly do serve a purpose within museums, for instance, or other companies and you know, businesses seem to like chatbots too, because 41% of IT professionals do expect AI to become critical to their business within the next three years.

G: That-, that is, that's very true. I find it interesting, you know, the perception between business versus customer when you're talking about chatbots. Uh, for me as a customer, chatbots in some cases can be quite helpful. I actually had one case where I was purchasing something online. It was a piece of furniture from my home and I actually was able to essentially buy my, the piece that I was looking for, by talking to the chat bot. It was able to direct me to the right thing and help set up something and then connected me with the human who helped me finish the process, which was interesting. But then I've also had cases where the chatbot just won't go away. [laughter] Which is, you know, it's nice to know that it's there, but sometimes I do know where I'm going if I've been on the site a couple different times. And when it won't go away, that can get, that can get a little annoying.

And, and what's interesting, too, is that Gil Press states in a recent Forbes article that 86% of people still prefer to interact with a human being instead of a robot. I personally am kind of similar in that sense. I just, I think when you're looking at it from an arts management perspective, how the different institutions are using chatbots is the most interesting bit for me. Personally, I wouldn't necessarily want to deal with the chat bot but seeing how institutions are actively able to implement these and do it effectively and increase [inaudible] interest, excuse me, from the people that they're serving is really intriguing.

A: Absolutely, maybe there's also an artistic aspect, like behind chatbots that can potentially help with that interaction? There was one article that I read online featuring an interview of Arash Rod and his chatbot designs. He's actually made them a little bit more human, not within the voice and the cadence itself, but rather in incorporating animated character designs as GIF responses or by having them take on a certain personality. So for example, like somebody like will say thank you to the chatbot, and there's an emotional like, GIF that gets sent their way and they take on a particular characters, say if it's from a sports chat or if it's from the video game chat, and this has seen very highly positive responses. So, potentially because it's one factor being that people like to see their favorite character, their favorite sports person.

G: Yeah. And thinking back to Dot from Akron Art Museum, the sparking conversation topics, I think, that pulls in the spontaneity piece that I feel a lot of individuals who are, you know, talking about their opinions on chat bots, or those kinds of bits of technology that arts or institutions are able to implement. I think having that spontaneity feels closer to having the human to human interaction, even when you know you're dealing with a robot or dealing with a piece of technology. So, I feel like understanding the comfort level is, is really important when you're considering implementing AI into your institution.

Yeah-, I…, chatbots they're, they're a fun thing. Because they can be used really effectively and then others, they're just, they're still very basic and still very automatic, and then they fall flat, you know, and there's a big difference between a chatbot being spontaneous or a chatbot potentially annoying or scaring away the users too.

A: But, yeah, sure.

Okay, so moving onward to our next topic. We're visiting a subject that isn't exactly new: hologram concerts. Now, ever since the release of the Whitney Houston hologram tour, fans have expressed a great disdain and disgust according to the article from Janice Williams of Newsweek. Now other hologram tours that have taken place, had great success and positive reviews and this includes Tupac, Frank Zappa, Amy Winehouse, and the legendary soprano Maria Callas. Of course, there have been a few negative comments here in and there on the capitalism of deceased artists, but with the recent rise of criticism on the Whitney Houston hologram tour, I'm starting to wonder, is there a real ethical issue with holding these hologram concerts?

G: So, when you're considering artists who have passed and artists who made a real impact on the people who followed them, I think there is an ethical issue here because when you're…, when you're talking about creating a tour of someone who did all this great work and, you know, had this massive fan base, and there were many different things that were specifically attributed to that artist for that reason. Um, under-, understanding the purpose behind doing a tour like this, I think is really crucial. And having a purpose that is meant to heighten the artists, from a managerial standpoint thinking, you know, if you're trying to extend the legacy or revitalize the legacy of an artist that you feel very strongly about, I think that might be a really, I don't want to say noble purpose, but a much more ethically sourced purpose to do a tour like this. Especially considering that a couple of these artists did pass in either violent or unfortunate ways. So, understanding the context of, of the artists specifically in relation to the audience as well as current day, I think is important. And if you can do that, which it seems like the Whitney Houston tour is attempting to do, I think that's where you start to kind of go on the good side of ethics maybe.

A: Well see, you know, it is interesting because I was wondering this a little bit too, and I did some research, and I was able to find a quote from Whitney's mom, Pat Houston, and her comments on the hologram tour. Now it's important because in order for these tours to take place, they must get permission from the estate to design the hologram.

G: Which is really crucial.

A: Absolutely. Yeah. And you know, like Pat Houston like she mentioned that the family was happened to her. She mentioned that the family was happy to work with base hologram and recreate an, quote unquote, authentic Whitney. She also commented that her fans deserve nothing less because Whitney gave nothing but her best. And, you know, it makes me wonder like, if Base Hologram has attempted to reach out to different estates for other artists as well, but maybe that question of if they've experienced a violent death, or if there's other controversies taking place, like maybe they've had quite a few estates denied the, that hologram

G: That’s possible. And I think there's also something too about when we had the first one, which I think was Tupac. I don't remember the year, but it's been several years since the initial development of this technology and the first tour using him or his hologram I should say. I think when it's something that's new, there's…, there's not a whole lot of conversation spinning about this because it is so new there. The questions haven't been asked yet. And I think now that we've progressed and have had several different artists who have had this done either for them or to them, depending on how you want to look at it, and these questions are starting to be asked now. So, I think it's just a result of time and of technology being used more, more often and taking advantage of being able to use these holograms to create these tours. I think that capitalistic bit is where people are starting to now see, oh, well, you know, if someone can just essentially bring my favorite singer back from the dead, like, what does that mean? And when you're considering it from an ethics, and again, a managerial standpoint, that's something that needs to be troubled a lot. And I don't know…, I think it could go either way, depending on how you contextualize the tour itself.

A: Yeah, it's interesting, because Base Hologram seems to have overwhelmingly great positive responses over different detractors or other negative responses before. And that's a quote directly from Brian Becker, the CEO of Base Hologram. And, you know, supposedly, maybe one positive way of looking at it is, let's say you wanted to see an artist like, back in the day, you know, you just were not able to, and unfortunately, they've since passed. But now suddenly, like there's this new opportunity to go and see them live or their hologram Live in Concert. Or maybe if you happen to have kids, and you're like, “Oh, you know, I saw Frank Zappa, you know, like, that was a really great concert, I wish you could have seen that”. And now again, like here's this opportunity suddenly rising. But then you also get a little bit into authenticity between the actual, honoring artists. Then-

G: Which I think is maybe the question there, of using tech to bring in a quote unquote, authentic artists back. I think that may be where some people are having some issues qualifying the two together in a way that makes sense in, in, in thinking of how we perceive arts at this point in time.

A: Yeah. And, you know, that's actually been a small complaint as well, a lot of viewers. A few viewers before have gone to see these concerts and have commented that they actually see the hologram of [inaudible] onstage and feel kind of creeped out by it because they know that it's not human, it's but it's just a little bit off, you know, similar to the chatbots from before.

G: Right.

A: And it's like that little bit of a difference that's very unsettling to them.

G: Yeah, I think that's something that arts managers constantly have to deal with the-, the sense of uncanniness that might arise from, whether it's the cadence that a chatbot is using or how it looks and how it moves. We as humans are very disturbed by [laughter] things that move and look like us, but we know are not. So, I think that's, that's something that-, that's something we're going to have to get used to as it becomes more heavily used within the art sphere and just across the world in general, doesn't even have to be the arts specifically. Um, but yeah, that's, I could see how that would that would creep some people out when you, you go, and I'm thinking too someone who's been to a previous performance when the singer was still alive, and then going back and seeing and knowing that it's not the actual person, that would be a bit offsetting for sure.

A: You know, it would be interesting, come to think of it, to see if somebody had created a combination of hologram art and artificial intelligence as well? Because imagine traveling to a museum to go see, I'm going to use the Mona Lisa as an example, and maybe there's a feature of the painting as well, where you have the opportunity to see a hologram of Mona Lisa and then ask her some questions and then get some responses back. And-, as if it were actually like Mona Lisa speaking, although like maybe that might also come across as creepy. But that could also be very informative as well and a pretty decent educational rate opportunity perhaps.

G: Right. Yeah. And I think it comes down to how, how the exhibit or how the event is being presented, you know, what's the purpose behind it? Is it educational? Is it you know, to revitalize the audience or give something back to the audience? I think it all goes back to the purpose behind it and what spurring the decision to use this technology for sure.

A: Well, hey, Grace, thank you so much for joining me for this talk today.

Now coming up on the AMT Lab website: So, speaking of artificial intelligence, for Whitepaper Wednesday we have “The Downsides and Benefits to Incorporating AI in Museums” brought to you by contributor Jiashun Fang. This article also briefly discusses chatbots, but additional topics to consider include automation and AI roles to enhance the museum experience. After that, we have a book review coming out by contributor Jamila Jordan. This book is Marcus du Sautoy’s The Creativity Code: Art and Innovation in the age of AI, and it offers the following question: Can AI complete creative tasks better than humans? And finally, part two of the inclusivity in video games article from the Master of Entertainment Industry Management students and capstone project will be released. This article will go over survey results regarding habits, perceptions, and community in the video game industry. All these articles can be found at www.amt-lab.org.

[Outro music]

-----------