Introduction - Cristina Oliva Patrick
🫛 This edition of AI, Software, and Wetware features an audio interview on ethical AI with Cristina Oliva Patrick, a 🇺🇸 USA-based 🇮🇹 Italian advocate for responsible AI and the author of The Responsible AI Brief. We discuss:
Living the contradiction of advocating for responsible AI while also using some AI
Participating in the free Lovable event for International Women’s Day to build a myth-buster app and a self-assessment app for her Four Pillars of Responsible AI Governance
How LLMs handle different slangs in Spanish and other high-resource languages she uses
Opt-out experiences with Google Gemini, LinkedIn, and other sites
Her upcoming chapter in Volume 2 of the She Writes AI Everywhere book series, covering low-resource languages in LLMs and the inequity at the heart of AI development
and more. Check it out, and let us know what you think!
This post is part of our AI6P interview series on “AI, Software, and Wetware”. Our guests share their experiences with using AI, and how they feel about AI using their data and works.
This interview is available as an audio recording (embedded here in the post, and later in our AI6P external podcasts). This post includes the full, human-edited transcript. (If it doesn’t fit in your email client, click HERE to read the whole post online.)
Note: In this article series, “AI” means artificial intelligence and spans classical statistical methods, data analytics, machine learning, generative AI, and other non-generative AI. See this Glossary and “AI Fundamentals #01: What is Artificial Intelligence?” for reference.
Interview - Cristina Oliva Patrick
Karen: I’m delighted to welcome Cristina Oliva Patrick as my guest for “AI, Software, and Wetware”. Cristina, thank you so much for joining me on this interview! Please tell us about yourself, who you are, and what you do.
Cristina: Karen, thank you so much for having me. I’m really excited about this. This is my first time on a podcast, and yours couldn’t be a better place to start, so thank you.
Hello everyone. I’m Cristina Oliva Patrick. I’m originally from Rome, Italy, but I’ve been living in the USA for about 12 years now. My background isn’t technical, but it is in the humanities, and I’m currently an equal employment opportunity specialist. I consider myself a community builder. I publish a weekly newsletter called The Responsible AI Brief on Substack where I explore the intersection of ethical governance and AI. And I also co-host AI safety meetups in Oregon and Washington. And I do that to help to bridge the gap between tech and the people it impacts.
Karen: Great. Thank you for that introduction. That’s a really important area that you’re working in – bridging tech and people. I’m happy that you’re doing that and I’m glad that we’ve connected. So tell us a little bit about your level of experience with AI, machine learning and analytics. You said you’re not technical, but have you used it professionally or personally, or have you ever studied the technologies?
Cristina: Definitely, definitely. Thank you, Karen. Yes. I started studying LLMs and AI in 2022, especially when the whole AI boom started to be very mainstream and everyone was using AI tools or chatbots. I focused on ethics and safety, of course, because it’s embedded in me and it’s my background.
I use large language models like Gemini and Claude to help me summarize research that I do. Complex papers, also brainstorm structures as well. Kind of a companion, almost – a companion – using it not as a friend or a mental health tool, but more of a work buddy, I guess, a colleague.
And also, yes, I definitely use it for identifying mistakes and errors in my writing, of course, as well. I mean, in the end, English is my second language. But most importantly I double check everything that these chat bots generate. I think that’s the most important thing for me.
I also use them to create images, but it’s something that I’m currently trying to think differently about, because I would like to find a more creative way to do it. I’m not a drawer really, but who knows? Stay tuned. I use the AI to create images for my newsletter, but I think there may be better ways.
Karen: Is that mostly because you’re not satisfied with the images that you get? Or do you have ethical concerns about the tools? Or both, or something else?
Cristina: Yeah, I think they actually look really good and they’re interesting. Even the process to create them with natural language is kind of fun. But they do require a lot of iterations. So I do have ethical concerns about sustainability. And also I feel like I live this contradiction of advocating for responsible AI, but also using AI. So I wonder if, at least for my job, maybe creating images isn’t really something that I need to do. Maybe it’s something I could avoid.
Karen: Yeah, I have some references. I don’t generate AI images for any of my work, and I have been building up this list of other sources of images that are truly public domain, not just publicly available, but that are free to use even commercially. And there’s some great sources of images. So if you find a source that you like, maybe that could just become part of your branding for your newsletter. So that’s always an option.
Cristina: Oh, thank you, Karen. I would love to see your resources. Thank you.
Karen: Yeah, happy to share that. [link to ‘ethical shoestrings’ resource list]
Can you share a specific story with us on how you’ve used a tool that has AI or machine learning features? I’m curious about how the AI features of the tool worked for you or didn’t. Basically, what went well and what didn’t go so well?
Cristina: Yeah, so I’ve recently used Lovable, which is an AI app builder, I guess for what we call now vibe coding. So creating something through natural language. And I created some short educational interactive websites, I guess we could call them. Write, like, a one-pager sort of thing. I did a myth-buster around women in AI. This was during International Women’s Day, and this happened because that day Lovable made their platform free. So I decided it was the perfect time to practice a little bit.
And I’ve been definitely inspired by a lot of women here on Substack. Not only you, Karen, but Elena | AI Product Leader, Code Like A Girl, Karen Spinner, Daria Cupareanu, many others, and sorry if I mispronounce your last names? My apologies. So yes, I was very inspired by all of them and more to try to see if I could create something or how to really look into these platforms to see how they work.
And I have to say, the platform, it’s very user-friendly. There are others. But yeah, especially for a simple project like that one that I did, it was really easy to use for a person like me who isn’t technical or doesn’t know anything about coding. And that kind of project doesn’t require any data from users. And so I felt it was easy and secure for me to do that.
Karen: Thanks for sharing that. Can you share a specific example of something that went well and something that didn’t go so well? For instance, when you tried the free day on Lovable, were you able to get something useful working within that free day?
Cristina: Yes, I did, that myth-buster. Again, like an educational interactive website. I made a couple other things that I’ll be sharing as well. One was around AI governance framework. So I was able to make a few things that I think are, I would say, 90% complete. It was easier because obviously everything was free, so you didn’t have a limited iteration limit or tokens or something like that. And I don’t have an API, so I had to keep it really simple, and so it was pretty easy. For me, at least, everything worked pretty well. Obviously, they required quite a few iterations, right? So you’re in the platform for some time trying to have this website looking like you want it, or app.
But it was a good brainstorming and practice day. So I feel like I have a better understanding of vibe coding and what people are talking about a little bit. And I think, for me to talk about AI and really understand it and understanding the risks as well, I need to use the tools and learn how they work.
Karen: You also mentioned using AI tools for summarization or for translation. Have you run into any cases that it did not handle well?
Cristina: Yeah, I think one of the biggest problems is that these tools can produce a summary or transition that sounds perfectly professional and confident, but unfortunately they do contain mistakes or false information, false citations that we call so-called hallucinations, these mistakes.
So that’s where we really need to double-check whatever outputs we get, especially for doing something regarding a summary that has a lot of research, a lot of numbers. I think these tools are still not very good at numbers or stats and especially citations as well. Many times, I see they cite a website or a study, and it’s not even true. But I am noticing more accuracy, for instance, in the cultural nuances of Spanish, Italian, English, which are the languages that I speak or I study. But of course these languages are mainstream languages, so I think that’s easier for these tools, obviously, to give us better translations in a language such as English.
Karen: Have you noticed that these translations are improving since you started using these tools more than three years ago?
Cristina: Definitely. Definitely, yes. Especially in the cultural nuances. Like, if I ask for some slangs of specific regions, even with Spanish, right? There’s a lot of different Spanish. So yes, I think they’ve gotten better. But again, I only can speak for this kind of more high-resource languages. So I don’t know for low-resource languages and dialects. That’s something actually that I want to study more.
Karen: Okay. Yeah, that’s a really great point that tools are trained disproportionately on English languages, Western cultures and such, so they tend to be less accurate on other languages and fitting into the assumptions and mores of other cultures.
Cristina: Exactly.
Karen: Yeah. Yeah. So you use AI tools for some things. Are there any things that you would avoid using AI tools for? If so, can you share an example of when, and why you would choose not to use AI for that?
Cristina: Yes, definitely. So I avoid asking for medical information, although it’s very tempting, or using it with private information. I always edit that information, or just choose not to input it at all. And like I said, I also would like to avoid using it to create images. So I’m excited to learn more about that with you as well, Karen. But yeah, it’s kind of an interesting duality because I do want to learn the tools. I don’t hate technology, and I like technology, but I also want to understand these tools and especially to be able to write around AI governance and policy. But I definitely avoid using it for medical information and I guess probably for mental health. I do not talk to the chatbot as a therapist.
Karen: Yeah, that’s very fair. A lot of people share your concerns about confidentiality and also about the duality and the contradictions, and you mentioned sustainability concerns before. One concern is about protecting your data. It kind of leads us to thinking about where these systems get the data that they use for training the tools. And a lot of times they will take data that people put into online systems or publish online, like our writing. I’m curious, have you opted out on Substack from allowing AI tools to use your writing in your newsletter?
Cristina: Yes, I did opt out. And it’s interesting that you mention it because yeah, nowadays we have to opt out, right? We are automatically opted in, so we have to make a conscious choice. We have to know it. First off, maybe other people that aren’t in this area, in this field, they don’t even know that. Same with Gmail and LinkedIn; you have to opt out. So yes, I did that.
I think right now we’re in a wild west phase where data is really treated as a raw, natural resource to be extracted. And especially, I’ve noticed that in the US, we don’t have a lot of protections around that. So I think right now, these big techs are just doing whatever they please with our data. So unfortunately the responsibility is sort of up to us right now to have at least a little bit of a shield, I guess, for ourselves, to be able to make sure our data are our own. But it is really difficult. It’s really difficult. It would be good. What I would advocate for is that we need our data to be our own. Our biometrics, our image, our art, our books, texts, articles, they should be ours, right? We should be the owners. But at this time in the US, I don’t see that happening.
Karen: Yeah. A lot of people advocate for what the Cultural Intellectual Property Rights Initiative calls the 3C’s rule, which is that there should be Consent, Credit, and Compensation. And that goes to what you’re saying about the ownership, that we should have the right to choose whether our work gets used, and we should be credited if it’s used, and compensated. It sounds like you agree with that, but also you’re acknowledging that that isn’t reality right now for most of the AI tools that people commonly use.
Cristina: Yeah. Yeah. It doesn’t seem like it, from what I read, at least in this part of the world. But yes, I mean, even something so personal, like our image, and yes, our biometrics, I don’t think. They’re out there, right? And there are some apps that explicitly say “If you’re posting on this app or sharing your videos or voice, we own it. We’re going to use it.” Maybe in ads or who knows, and you’ll not even be compensated.
Karen: Yeah. And in many cases, not even consulted.
Cristina: Right, exactly. Exactly.
Karen: Yes. Yes, exactly. So as someone who uses these tools, do you feel like any of the tool providers have been transparent with you about what they will do with your data and about where they got the data that they used for training the tools?
Cristina: . I don’t think so. I think transparency is probably one of the biggest problems right now. I mean, most providers offer models that are black-box, right? So we don’t really know exactly how what we input is going to be used, what training data they’ve used, even. So for me to see this is very frustrating. And as a person who works in policy and AI governance, I would like to be able to do more, and do something, especially for organizations that might use third party AI tools. But definitely I think transparency is a big problem right now.
Karen: Yeah. Often they’re talking about data that’s publicly available, and they’re conflating that with public domain, which is not at all the same thing. Just because someone publishes something online, that doesn’t mean it’s not copyrighted and it’s not their intellectual property that they own and should have the rights to control. So they are being disingenuous about that in a lot of cases.
Also interesting, I don’t know if you’ve heard this, but they’ve made some claims that we couldn’t possibly get good-performing models if we didn’t steal everything, basically. But that actually turns out not to be true. There is a project called Common Pile – I think it’s US and Canada researchers – and they built a model with only ethically-sourced, licensed data, and they got good results. The models performed well. So claiming they need to do that is a lie. [link]
Cristina: No, absolutely. Thank you for sharing that. Actually, I haven’t read that, so I’m going to go and investigate. Thank you.
Karen: Yeah, I think it’s interesting that they say they need our data, but it’s really, I think, more just a matter of, that’s the way the business models are set up right now. They aren’t being penalized enough for doing that to offset the money that they’re making from it.
Cristina: Exactly. They’re going to do whatever they want. Exactly.
Karen: Yeah. And they’re not currently being held accountable. As you pointed out, in the US, we don’t have good regulations that control what people can do with our data.
Cristina: Exactly. At least not yet. You know, I always have to keep my hopes up!
Karen: Yeah. Yeah. Just as consumers and members of the public, even if we aren’t trying to use AI tools, our personal data and content have probably been used by AI-based tools and systems. Do you know of any cases that you could share when your data has been used? (Obviously without disclosing any personal information.)
Cristina: Yeah, there’s a lot of content of mine online, so probably it’s been used without me knowing. I hope that my personal data aren’t really out there, PII [personally identifiable information], private information. But I’m sure that, for instance, probably my phone number is out there, right? That’s why I receive a lot of spam calls.
Actually, I think I have two examples. I sort of mentioned it already, but Gmail for instance, you have to know that they will be training. This was pretty recent, right? As well, they decided to add their AI Gemini to Gmail. And so you have to know that you had to opt out. You were automatically opted in for your data to train Gemini. So you had to know, first off, and you had to make a conscious decision to opt out.
And TikTok, for instance, recently changed their privacy policy. So really, whatever you upload there, it belongs to TikTok and not to you. So yeah, these platforms are not an opt-in. They’re mostly an opt-out type of platform. Sometimes you don’t even have the choice, like with TikTok, if you want to use it, you have to pretty much give your image, voice and everything to them.
And so, yes, again, I feel like they’re shifting the burden on us, on the users and on the individual rather than the big corporation, which obviously I’m not. That I don’t agree with. I think the main responsibility should be on these corporations.
Karen: Yeah. So are there any social media platforms or such that you simply won’t use because you have concerns about their policies? I know some people, for instance, run their businesses through Instagram, and so when Meta started asserting more claim to their material, it became hard for them to say they would leave. I’ve just never used Instagram because I don’t care about it, and I don’t like Meta’s policies, and so I’m just not going to use it. I have a choice. But some people don’t. How have you approached using social media with that in mind?
Cristina: Yeah, no, same here. Like I said, I started using social media when I was pretty young and I didn’t have any idea of these topics, right? Data privacy. I had no idea. I just wanted to do what everyone was doing and share things with my friends. So I definitely have a lot of content out there. And I cannot scrape it away, remove it. You can still find it. But definitely, I don’t use TikTok, for instance. I’ve never used it. I’ve never felt comfortable – again, their privacy policies – I don’t know, wishy-washy, not very good.
But again, I use LinkedIn and there’s a lot of controversy with that as well. Also with the verification process that they use, you have to send in your id. How is that stored? They use a third party, so there’s a lot of controversy. I’m aware now, but it’s hard because I am online, right?
Karen: It’s not quite two years ago when LinkedIn had that fuss where they tried to assert the right to everything we’d ever put in the platform up to that date and said, “Well, you can opt out going forward if you want to.” That was not cool. There’s actually two ways you have to opt out. And I found out, not from LinkedIn, but from other people who were posting about it. So I opted out of those. [Readers: here’s how - 1, 2, and now 3; also check ads.] But they still claimed and took all of the data that I had in there prior to that point, which I wasn’t thrilled about. I’m finding I’m using LinkedIn less and less lately.
Cristina: Yeah, no, and they’re not very transparent, right? Like you said, you sort of have to figure out even how to opt out, right? You find it through other people posting the instructions or the news about it. “Hey, this change is coming to this platform; you’ve got to do this.” So it’s usually actually more users that are trying to help other users. It’s not really the corporation telling you.
Karen: The verification is especially tough. It’s one thing if you say, “Well, I just don’t care about getting that special check mark”. But in some cases, I know some people have lost access to their accounts, and they were forced to either verify, and provide all of that very personal identity information, or they would lose access to their account permanently. And that seems exploitative.
Cristina: Yeah, 100%. Yes.
Karen: Yeah. So it’s good that you’re cautious about those and aware of those, but yeah, it should be on them to tell us. We shouldn’t have to discover it by accident.
Cristina: Exactly. It should be on them not to use our data. But I almost want to say it’s too late because the whole internet is built on our data, and all the ads. So, yeah, I don’t know. I don’t know.
Karen: In fairness, this is not really a problem that’s new with AI. This problem with data brokers scraping and stealing and selling our data has been around for decades now, so it’s not new. But AI makes it worse in some ways. Because if some model gets trained on the data, even if we get it removed or corrected, it’s still embedded now in those models forever, and it’s really not going to come back out.
Cristina: Yeah, that is a good point. Yeah, it is definitely making it worse. Yeah.
Karen: Yeah. You mentioned Google and Gemini, so it sounds like they were at least relatively straightforward about informing you, even though we were opted in at first. I think I remember seeing a banner come up in Gmail about offering these overviews and you can opt out, and I immediately opted out.
Cristina: Yeah, exactly. Yeah, same here. I feel like it was a little bit, at least it made us aware, because they probably have to. But again, I also probably saw more information and articles coming from users, rather than Google. But yes, at least I saw that banner too, for a little bit, at least, I would say.
Karen: Yeah. It sounds like it’s usually coming as a surprise when the company finally acknowledges publicly that they’re doing this?
Cristina: Right. I mean, I knew with Gemini it was coming because they have so many products, right? And it was only a matter of time that they were going to embed Gemini in everything that they offer. So yeah, that wasn’t a surprise for sure.
Karen: So does your work in responsible AI tie into that? Do you help companies to behave better and more ethically in these regards? Is that part of what you do?
Cristina: I want to, yes. This is what I want to do. Exactly. Thank you for bringing it up. I’m still working on that. I’m still trying to pivot my career for sure, but that is something that I would want to focus on as well.
Karen: Are there any other times when you’ve opted out from some use of your data? One example would be travel, where they have this very small print on the TSA signs telling you that you can opt out. I’ve had some people say that they’ve done it successfully, and I’ve also had some people say that it was too much hassle or it didn’t feel safe to do. And I’m wondering what your thoughts are, your experiences?
Cristina: I agree 100%. Yeah, I tried a couple times to opt out from the TSA’s facial recognition. At the very beginning, I didn’t even see signs. Since I research a lot about these topics, I learned that it was totally voluntary. But they didn’t tell you that. They don’t tell you that at TSA, right? So I tried a couple times. Both times, the TSA officer wasn’t really thrilled, or it was a little bit of a hassle. Especially in the US, I think my information is everywhere, because I’m also an immigrant. So there’s a lot of documentation about me, let’s just say that. But also, especially I think for people who may not be citizens of that particular country, or immigrants, it might be really intimidating to say no to a TSA officer. So I can see how they definitely have some power there to get what they want.
Karen: Yeah. It is interesting. We both wear glasses. I just got new glasses recently when my prescription changed. And the company that I bought them from, Zenni Optical, they now offer this special lens tint that is supposed to protect you somewhat from facial recognition. Basically a tint disguises your eyes enough that facial recognition or infrared scanners can’t pick up the shape of your eyes. I didn’t get it, because mostly I just wear these glasses at home, so no company’s recognizing my face here. But I thought it was interesting that this is now a product offering from an eyeglass company to try to help people be less vulnerable to facial recognition techniques.
Cristina: Wow. I mean, there’s definitely a market. Do you think it works?
Karen: I don’t know. I’ll drop you the link to the company if you’re curious.
Cristina: Oh yeah.
[link: Zenni Optical ID-Guard; story about ID-Guard]
Karen: But I thought that was very interesting to see that they were offering that as a product. They claim it works. I didn’t get it. And it would be interesting, I suppose, to test it, but glasses are kind of expensive to test on.
Cristina: Wow. Interesting.
Karen: Yeah, anyways, thank you for sharing that story about facial recognition. And again, the concern there, I think is, it’s one thing if our stories or writing gets stolen, you know, that’s ephemeral. But our biometrics, our faces are our identities, our passports, those are extremely difficult to change if they get stolen. And so that’s a real concern.
Cristina: Yes, I agree: our biometrics, I think that’s probably one of the scariest things around AI, for sure.
Karen: Have you ever had any incidents where a company using your information has caused a specific issue for you, such as privacy invasion or phishing attacks, or has it cost you any money ever?
Cristina: Thankfully I haven’t experienced any issues with my personal data being misused, knock on wood, or it didn’t cost me any money. Again, fingers crossed.
However, I did have a close call with the phishing attempt. I’m kind of embarrassed, but at the same time, you know, it does happen. I almost fell for a fake interview invitation. It really seemed legitimate at first glance. I even had a little bit of a back and forth. We exchanged a couple of emails. Even the email address looked professional, the text as well. I don’t know why, but something didn’t sit right with me. So thankfully I had the idea to actually call the company directly. So I just search the company directly to get their actual real number. And so I was able to verify. And so it turns out that yes, it was a scam. I’m not sure how the scam was going to play out, to be honest; I don’t know. And I didn’t really give them my information. But I guess they had my email, of course, because of my resume, so they did have some information.
But yeah, the lesson, I think, is always double check, right? Especially when something feels off, but even if it looks real and legitimate on the surface. I guess the first time I responded to the email, I didn’t really think much, right? I didn’t really pause. But I think that pausing is really what we need to do.
Karen: Yeah, that’s a really good point. Especially if they think about the data that’s on your resume. It’s very tempting, especially in this job market for people to, “Oh, someone’s finally reaching out to me about a job!” and just respond to it. And sometimes some people have even more personal information on their resume. For instance, it used to be a common practice to put an address in, days ago. In some countries, they’ll put a date of birth on their resume, which floors me nowadays. But back then, that was considered a normal thing. So sometimes they’re just trying to get your resume to get your home address, to get other information that you may have on there, or to have other ways to maybe go after other people.
I actually saw something on LinkedIn a couple days ago. Someone was posting about the current job market and how tough it was to find a job. And someone commented on it and said something like, “Oh, there’s a great opportunity for you at my company, and here’s the email address for the recruiter.” And, you know, “You should get in touch.” But it wasn’t a company email address; it was like something@outlook.com. A legitimate recruiter is going to have a company email address. Seriously? It just smelled like a scam. This person was supposedly an executive director at the company, but they had no connections, and almost no history. It makes me angry to see people that would try to take advantage of people like that. That just doesn’t sit well with me.
Cristina: No, I agree. I mean, especially now, like you said, the job market is really tight right now. And a lot of people are looking for jobs. So yes, it’s really outraging that people are trying to take advantage of people’s vulnerabilities, right? But yeah, I think, again, pause. And probably calling a real person, not that the person that was scamming me wasn’t real, but calling to verify or meeting a person. I feel like that worked for me because it was a company that I applied for and also the email looked legitimate. It was really not that different from the actual company’s email. And so, again, thankfully I called and asked them about the person that was emailing me and they said, “No, we don’t have any idea.” And they also told me about their actual email. So yeah, it’s sad.
Karen: Yeah, unfortunately the fakes are getting better, and it’s more work for us to figure out if they are fakes or not. But yeah, so important now that we protect ourselves that way.
Cristina: Yes, I agree. Yeah.
Karen: So last question, and then we can talk about anything else that you want. We talked a little bit about companies misusing our data. And we’re seeing that public distrust of the AI and tech companies has been growing lately. It’s probably healthy distrust. What do you think is the most important thing that companies would need to do to earn and then to keep your trust, if that’s possible? And if you have specific ideas on how they could do that, if so?
Cristina: Yes, absolutely. I think first off, governance and guardrails should always be built in from the very beginning and not just as an afterthought.
And also transparency and accountability. I think that companies just need to move beyond that virtue signaling that a lot of companies do. Sometimes that happened even with diversity, equity, and inclusion. Some companies would just put a disclaimer over there. But I think you have to move beyond that and actually go toward verifiable accountability, actions and frameworks. I think it would mean being clear about data provenance.
Also, I like this a lot, the ‘right to be forgotten’ requests, right? And actively involving voices from every background – not just engineers, but also other backgrounds, also from the humanities as well, in the development phase, right?
I think that was it.
Karen: That’s a good answer. I appreciate you sharing that. I’m so glad you made time for this interview with me today, Cristina. Thank you. Is there anything else that you would like to share with our audience?
Cristina: Thank you, Karen. Thank you again for inviting me to speak with you and to be interviewed by you. Thank you so much.
I created the four pillars of executive AI governance – actually goes well with what we were just talking about. And we can share the link as well in the article that’ll go out with this interview. I wanted to provide a practical framework for leaders to balance ROI, so economic gains, with responsibility. And so the pillars for me are accountability, transparency, fairness, oversight. And I also created a 90-day action plan to help leaders. I would love to talk more about this, but there is a full article on this, and we can add the link if you don’t mind, Karen.
But I love to talk more about this if people have questions, if anyone might be able to use it. I would love to learn more about how they’re using it, their questions as well, and feedback.
Karen: Yeah, and we’ll definitely include the link to your Lovable app and the demonstration for that. I’m really curious to learn more about it. Oversight, I think, is so interesting because one thing that we see with humans in the loop on AI tools oftentimes is that people tend to trust the tool too much, and human review can become … not meaningful. It’s more like a rubber stamp. To have true oversight can be a challenge. So I’m really curious to learn more about how you help leaders handle that aspect of their responsibilities.
Cristina: Thank you, Karen, for bringing that up. I definitely speak a lot more about it in the article. This was a collaboration with @Joel Salinas. I am going to write a follow up article and it’ll go out this week where I’ll talk more about this framework.
Karen: Okay. That’s awesome. Thank you Cristina for sharing all this information and we’ll definitely get those links to your articles and your newsletter in here.
I’m excited to share with our audience that Cristina has graciously agreed to contribute a chapter to Volume 2 of AI Everywhere in the SheWritesAI collaborative books. Cristina, would you like to say a few words about your chapter?
Cristina: Thank you, Karen. Saving the best for last. I love it. I’m super excited about this. And yes, something that I mentioned a little bit earlier, but the chapter is going to be about low-resource languages and how large language models reshape how the world communicates, and how these low-resource languages have the risk of being completely locked out. I will talk about the inequity at the heart of AI development where English and a handful of dominant languages receive huge investment and thousands of other languages may vanish from our digital futures.
Karen: Yeah, it’s a great topic and it’s one of the aspects of bias in large language models that is sometimes not easy to see. And I think it’s great that you’re shining some light on it, not just in this chapter, but also in your Substack newsletter, your Responsible AI Brief. And I hope people will check that out and stay tuned for your book chapter. It’s going to be awesome. I can already tell.
Cristina: Thank you. Appreciate you Karen, and thank you for all you do for the community in AI. You do so much. I really appreciate all your work. I know it takes a lot of work.
Karen: It does. And I appreciate you doing the work that you’ve done earlier this year for organizing our first networking event. That was a lot of fun. I’ve had some people ask if we’re going to have more of those. So hopefully we’ll have a few more opportunities to do that this year.
Cristina: We definitely need to do more. Karen, I appreciate it. We need to bring the community together more often.
Karen: Thank you so much, Cristina. This was great, and have a great rest of your day!
Cristina: Thank you so much, Karen. Yes. Talk to you soon. Have a great day.
Interview References and Links
Cristina’s Four Pillars of Executive AI Governance: Article, App
Cristina Oliva Patrick on LinkedIn
Cristina on Substack (The Responsible AI Brief)
About this interview series and newsletter
This post is part of our AI6P interview series on “AI, Software, and Wetware”. It showcases how real people around the world are using their wetware (brains and human intelligence) with AI-based software tools, or are being affected by AI.
And we’re all being affected by AI nowadays in our daily lives, perhaps more than we realize. For some examples, see post “But I Don’t Use AI”:
We want to hear from a diverse pool of people worldwide in a variety of roles. (No technical experience with AI is required.) If you’re interested in being a featured interview guest, anonymous or with credit, please check out our guest FAQ and get in touch!
6 'P's in AI Pods (AI6P) is a 100% human-authored, 100% reader-supported publication. (No ads, no affiliate links, no paywalls on new posts). All new posts are FREE to read and listen to. To automatically receive new AI6P posts and support our work, consider becoming a subscriber:
Series Credits and References
Disclaimer: This content is for informational purposes only and does not and should not be considered professional advice. Information is believed to be current at the time of publication but may become outdated. Please verify details before relying on it. All works, downloads, and services provided through 6 'P's in AI Pods (AI6P) publication are subject to the Publisher Terms available here. By using this content you agree to the Publisher Terms.
Audio Sound Effect from Pixabay
Microphone photo by Michal Czyz on Unsplash (contact Michal Czyz on LinkedIn)
Credit to CIPRI (Cultural Intellectual Property Rights Initiative®) for their “3Cs' Rule: Consent. Credit. Compensation©.”
Credit to Beth Spencer for the “Created With Human Intelligence” badge we use to reflect our commitment that content in these interviews will be human-created:
If you enjoyed this interview, my guest and I would love to have your support via a heart, share, restack, or Note! (One-time tips or voluntary donations via paid subscription are always welcome and appreciated, too 😊)
