AMT Lab @ CMU

View Original

Let's Talk: Impact of Covid-19 Vaccine News, Racism in AI, and Tech Resolutions for 2021

In this month’s Let’s Talk episode, Angela Johnson and B Crittenden chat about how news of the Covid-19 vaccine is affecting attendance at arts institutions, as well as racial biases in AI implementation in museums. They also share their technology resolutions for 2021. Do you have any tech resolutions? Let us know in the comments!

See this content in the original post


[Musical intro in background]

Angela: Welcome to the Let's Talk series of Tech in the Arts, the podcast for the Arts Management and Technology Laboratory. My name is Angela Johnson, the Podcast Producer.  

 B: And I'm B Crittenden, the Technology and Interactive Content Manager.  

Angela: Each month, we review trending stories and discussions with topics such as streaming, artificial intelligence, marketing, social media, inclusion, fundraising, and much more. Our goal is to exchange ideas, bring awareness, and stay on top of the trends. In this month's episode, we will discuss how news of the Covid-19 vaccine is affecting attendance at arts institutions, as well as AI implementation in museums. Also, B and I will discuss our tech resolutions for 2021. 

B: One of the main hot topics from the last month or so is the emergence of the vaccine beyond it just being a hypothetical. That is obviously huge news here in the United States. So Colleen Dilenschneider, who's an arts marketer, and who runs the blog, Know Your Own Bone, she has been compiling some data and doing an analysis on how the vaccination news and the rollout has changed plans for audiences when it comes to visiting cultural entities in the upcoming year. The measure she's really focusing in on is intent to visit, which rather than just like an interest in visiting a performance or a cultural space, intent tells when someone is actively planning to attend to that event or space, and it tells us how the current environment is impacting future plans. She has been gathering this information regarding intent to visit but really focusing in on it since March when the pandemic really started taking over in the United States. They saw a pretty significant change in the United States when it came to intentions to visit cultural entities. About a month ago, when the vaccination news started to hit. They measure intent to visit for 84 cultural entities in the US. And respondents quantify their intent by ranking their intention between one, which would be like no intention to visit, and 100, which is basically that person is definitely going to go to an event. In September, October, and November 2019, they were able to kind of set the standards for what audiences intentions for visiting were within three months within six months and within one year, so they feel that they could really anticipate what 2020 would have been had there been no pandemic. Vaccination news has significantly impacted these intentions to visit in 2021. We saw increases in intentions to visit within six months, and by the end of the year. So that means that people are planning to attend cultural entities, cultural events, more in 2021, compared to the same exact period in 2019. So it's pretty staggering, that spike, because I think it reflects people's intentions to get out and kind of make up for lost time. And Colleen, in her analysis, concludes that people are likely to be deferring their visits until after they get vaccinated, because there was a dramatic decrease in intentions to visit in three months, as of the end of November, reflecting that people are becoming aware that they likely won't be vaccinated within three months, but it's more likely that that they will be between six months in a year.  

Angela: One part of the article that felt very sad, but I recognized it a lot myself, it was the section called “The Sad Romance of the Moving Six Months Mark”. It reminded me of this whole pandemic and every time we're all like, okay, in six months, though, things will be back to normal, and we can go to museums and go to the theater and everything will be fine. And then six months go by and we're like okay, but six months from now. And now that there's a vaccine. We're saying six months, but I think we all kind of really mean it or there's a reason for us to hope, and I think that's really important because I don't want to have to move my six month mark again.  

B: No, I'm glad you brought that up because I love the way that she framed that. And I also think that that feeling that you just described of having a reliable six month mark is reflected strongly in her analysis and in the spike you see of intents to visit for the six months and one year mark. And it also just reading this, kind of like you described, I was kind of reliving the experience of like being in March and saying, well in three months, and then three months rolls around; well, by September, surely...  

Angela: There are so many hard things about the pandemic, but one of the hardest is just all of the uncertainty. Like we just don't know how long this is going to last and just giving us something to say: okay, here's when things will start to change. Here's when things will start to get better. Obviously, the vaccine's not going to fix everything. We still need social distance, and we're masks, but it gives us hope, and we need that. And I miss museums.  

B: I know. She's also been doing an analysis around what would make audiences feel safe and comfortable going to museums, cultural gathering places, concerts. And the number one factor that will make audiences feel safe again, both in October and the beginning of December, was mandatory face coverings. Coming in second was the availability of the Coronavirus vaccine. So yes, the news of the vaccine is very exciting. And it's giving people hope and intentions, perhaps planning for something and feeling sort of set in that. But, yeah, it's not a quick fix. And the face coverings will hopefully be sticking around as well as other social distancing measures.  

Angela: That there's a spike higher for 2021, than it was in 2019, I think really reflects how much people are missing arts institutions, but hopefully is going to help arts institutions recover from this year, because people realize the value that they were providing to society now that they can't enjoy them as easily.

[Musical interlude]

Angela: So, last month, we were talking a little bit about AI and the benefits that AI can have for arts institutions. And I thought that it would be interesting to talk a little bit about some ways that AI can be a little bit more problematic. There's an article from the American Alliance of Museums. And this is talking a lot about, for finding data, and for researching, and how AI can sometimes have some problematic and racist tendencies. And the quote is that "the most classic example of racism in AI, is when machine vision is used to make assessments about humans. Machines replicate the same poor judgments about people of color as humans, because the algorithms are built without scrutinizing the engineers' inherent biases." And that relates to museums because museum professionals - like, they outsource the work, that data can often be biased. And it's not really the machines that are biased. It's people that are programming the machines. And I think a lot of times we forget these machines are not objective. They're programmed by people and people have biases, people are flawed. And we need to be examining that. And just kind of acknowledging that. There's this really cool art project that was created by artist, Trevor Paglen, and AI researcher, Kate Crawford, which shows a lot of the biases in AI technology. It was called ImageNet Roulette, and it reveals how little-explored classification methods are yielding racist, misogynistic, and cruel results. It's really cool. It's a little scary. But basically, people upload pictures of themselves and the AI like sorts them. They noticed that people of color were generally either defined just by their race, or in some negative way, like criminal, but white people were often just defined by their career, like a doctor or news broadcaster. But if it was a black person, they would say, this person's black, or this is a criminal, and a lot of people started realizing that and they were like, wait, this is really racist. And that was the point of the art project: to show that this is what AI is doing. And we need to be aware of AI classification tools, because this is a very common thing within that technology.  

B: Yeah, it isn't an objective solution. It's not like the algorithms are just coming from on high or anything, you know. 

Angela: Like, we just don't think about it because we're like, ‘but it's a computer. A computer can't be racist'.’ And it's true: computers aren't racist. It's not even so much that these computer programmers are racist. They just they have a lot of inherent biases and a lot of computer programmers are white. And so they're just not thinking about it and they're not like programming against it. We need like anti racist computers, is what we need.  

B: I wanted to share a bit from a report released by the AI Now Institute in May of 2019, which highlighted diversity issues among AI engineers. By the way, the AI Now Institute at NYU is an interdisciplinary institute focused on researching the social implications of artificial intelligence. And the report is called "Discriminating Systems: Gender, Race, and Power in AI", and is based on a year-long pilot study. Basically, in terms of racial discrimination in AI, some of its research includes the findings that black workers are substantially underrepresented in the AI sector. Only 2.5% of Google's workforce is black, and Facebook and Microsoft are each at 4%. And then, Latinx workers make up about 3.6% of Google's workforce and about 5% at Facebook and 6% at Microsoft. Computer science, as an industry, is experiencing a historic low in terms of diversity. So the proportion of bachelor's degrees in engineering awarded to black women declined 11% from 2000 to 2015. And these numbers are not reflective of STEM as a whole, and that in fields other than computer science and AI, racial diversity has actually shown an improvement. It also addresses pipeline studies that have served to sort of assess the flow of diverse candidates for jobs from school into industries, and that there has been no significant progress in diversity in the AI industry from these, and I quote, "The focus of the pipeline has not addressed deeper issues with workplace culture, power asymmetries, harassment, exclusionary hiring practices, unfair compensation, and tokenization that are causing people to leave or avoid working in the AI sector altogether." So the report did provide some recommendations for increasing diversity in the AI workforce and the bias in the algorithms themselves: it talks about transparency around compensation, harassment, and discrimination, as well as a commitment to increase the number of people of color and other underrepresented groups at senior leadership levels of AI companies across all departments. It also recommends transparency around AI systems themselves and the need for more rigorous testing across the entire lifecycle of AI systems to ensure that there is constant monitoring for checking bias or discrimination within the systems as well as risk assessments to evaluate whether certain systems should even be designed at all, which I think is really important. So much of this research does not explicitly refer to the use of AI in museums or the arts. But it's absolutely relevant and definitely applies to this context, as well. I mean, many museums and other organizations are using the products designed by the companies cited in studies like this, such as Microsoft, Google, Facebook, etc.  

Angela: Yeah, adopting new technologies, I think, is really important for advancement in the arts and for just keeping the arts relevant. But also we should understand what technology we're adopting and who is creating that technology and how it works. 

B: Shout out to the author of the American Alliance of Museums article you referenced at the beginning of this conversation. Her name is Seema Rao. The article is called “Museums and AI: Is There a Ghost in the Machine”. It's one of three articles she wrote to cover the Museum Computer Network convening in Miami in 2019. I encourage anyone listening to go check out her blog and her Twitter because she's a good resource for conversations around museum tech, as well as user experience. I was not familiar with her before this, but I'm now following her on Twitter. And I've bookmarked her blog.  

 Angela: And I love the tweet where she says, "AI is not a sentient being. It's evil spreadsheets." Because that feels, uh, real. 

B: Yeah, I mean, as you initially said, the AI systems that are using physical appearance as a proxy for character are deeply problematic. And such systems are really replicating these biases in ways that can just like further deepen inequality. If you're interested in reading more about bias in AI, there are some other resources on AMT Lab. One of our contributors, Lauren Cornwell, wrote a piece called, "Considering Biases in AI and the Role of the Arts," which was published in March of 2019. And our podcast predecessors actually included a short conversation in their "Let's Talk" episode, which was published in November 2019. So they can be found on AMT Lab's website

[Musical interlude]

Angela: This has not exactly been the year for New Year's resolutions. I don't generally make New Year's resolutions, but even if I had, it's not like I would have gotten any of them done this year, because, you know, this year is kind of a wash. 2021, and vaccine news, and politics being what they are...2021, it seems like the time for hope and to try new things and change your habits and have resolutions. Do you have any, B?  

B: I have two tech-related resolutions. One kind of comes from the fact that I've been doing some research over the past semester on streaming, and this research has led me to reading and thinking more about just how streaming has changed how we're consuming media. And as someone who uses Spotify frequently—and my Spotify Wrapped analytics really hammered that one in for me—while I plan to continue to use it, I do recognize that it is a very poor generator of revenue, especially for emerging musicians. Basically, anyone not in the top 1% of streams on Spotify. If it wasn't clear to you already, the last year has made it very clear that artists rely heavily on income from touring or live performances, and then from selling copies of their work. So yeah, I intend to invest a little bit more in trying to support those creators and artists, especially those that are emerging by either subscribing to their Patreon, or buying copies of their work. Because I've been feeling very guilty about that. Yeah, just that like year-end Spotify analytics, I was like, oof.  

Angela: Oh, yeah, I know. I don't like to use Spotify, because I know that they don't pay artists really, at all. And I like to support artists. I just don't have a lot of money and it's cheap, is the problem. It's the like, you know, the sin of convenience that is like slowly killing us all. And it's the reason we all still use Amazon. 

B : Right, well, it's kind of like, I've invested in a monthly subscription for Spotify over the past years, and therefore haven't been like acquiring my own library of music. So if I were to go all of a sudden, like, buy all the music, I can't go just like buy all the music I want to listen to. So yeah, I mean, on the bright side, there are some solutions that have been presented: there is a music streaming company in China called Tencent Music. And this is a possible model that companies or a company in the United States could adopt, in which they facilitate micropayments from listeners to help compensate artists on their platform, in addition to the streaming royalty, which typically offer a pretty negligible amount of revenue for those musicians. And these micropayments can either be donations or given in exchange for virtual goods. And it can really help that like one-on-one transaction between the user and the artists on the platform, can kind of help bolster that artist-audience exchange. This platform kind of has elements of Patreon but those elements are integrated with the streaming itself as opposed to on on an entirely different platform. So maybe that's in the future, but for now, I want to be a little more thoughtful about how I'm consuming my media.

My second resolution is uh, this fall for the first time I have been introduced to the beauty of project management software and project management tools. And full disclosure: I haven't used any others—I know some people have pretty like intense loyalty to certain tools—but I've been using Asana this semester and I've really enjoyed it. And it's usually been for projects or classes, but I think I want to use it for like my own personal tracking of things for the next year. What about you, Angela, what are your tech- do you have a tech resolution? 

Angela: Um, I have a couple. One is like serious and then the other two are just like, fun ones for me, because it's been a hard year. And I'm not going to expect too much of myself. But my first one is that I am not good at budgeting, and I need to be better because one day I'm going to have expenses again because I won't live at home. And so I've been researching different like budgeting tools and stuff online. And one of them that I think is cool is called YNAB, which stands for You Need A Budget. And that one's really cool, although you do have to pay for it. There's another one which is free, which is called Personal Capital. And that one's actually pretty nice. And you can like link it to your banking and investment. They have a cash flow tool. So you can like see what you're spending your money on, which I need.  

B: Horrifying, but useful.  

Angela: I know. It's necessary. Um, but uh, yeah, so I'm definitely going to try to be better about budgeting in this next year. So I have a couple of more fun ones, though. One of them is that the library has this streaming service called Kanopy—with a K instead of a C—and they have a lot of documentaries and foreign films, and it's completely free. And I'm going to try to watch that more. Because I think it's great. I mean, they have this really great documentary on the New York Public Library. And I'm a nerd, so I just like libraries. And I'm like, oh, that's exciting! But they also have a lot of really great films that just aren't on Netflix, basically. And it really has more of like an arthouse vibe, and I like that a lot. So I'm going to try to watch for those. My my third one is very, very unrealistic. But you know, it's a resolution, it's fine. But basically, I've been watching this K Drama on Netflix, and they had these these like foldable smartphones. And they're so cool. I didn't know, like, I knew they existed. But I didn't know they were so cool. And they like fold, and then you unfold it and the screen is completely seamless. And I want one. They're really expensive. I'm not going to get one because I'm going to be better about budgeting. One day, I'm going to get a foldable smartphone. And so that is on my list as a goal.  

B: Angela, I believe in you on this one. I think- I'm really excited for you. 

Angela: Do you have any tech resolutions? Let us know in the comments. Happy Holidays. Thanks for listening to the AMT Lab podcast. Don't forget to subscribe and leave a comment. If you would like to learn more, go to amt-lab.org; that is A M T dash L A B .org, or you can email us at amtlabcmu@gmail.com. You can also follow us on Twitter at Tech in the Arts or on Instagram, Facebook or LinkedIn at Arts Management and Technology Lab. You can find the resources that we referenced today in the show notes. Thanks for listening. See you next time. 

[Musical outro]