
#techlondon · 2023-07-12 · 22 min
Key moments - from our scoring
Substance score
35 / 100
Five dimensions, 20 points each
Connected Women pivoted from a job-matching platform during the pandemic to focus on data labeling and annotation work - a critical but often invisible component of AI training. Gina Romero explains that AI tools like ChatGPT require millions of tagged images and text datasets to learn, and this labeling work is increasingly outsourced to the Global South where costs are lower. The episode explores the moral complexity: while companies benefit from cheap labor, workers in the Philippines can earn sustainable income without leaving their families. Romero discusses the pitfalls - workers in exploitative conditions, inadequate wages, and mental health risks - and how Connected Women mitigates them by paying Metro Manila minimum wage regardless of location, vetting projects for wellness impact, and partnering with UN Women. For B2B operators, this episode illuminates the hidden infrastructure behind ChatGPT and other AI systems, the geopolitical dimensions of AI development, and how to engage ethically with outsourced talent in emerging markets.
Data annotation is tagging millions of images with text labels or boxing objects to teach computers to recognize patterns - a process that ChatGPT and other AI tools require at massive scale to learn language and vision. Companies outsource this work to regions with lower labor costs, like the Philippines, to manage expenses while training sophisticated models.
Connected Women pays workers the Metro Manila minimum wage regardless of location, carefully selects projects that won't harm mental health, teaches workers to take regular breaks and practice good work habits, and has been vetted by UN Women for women's empowerment standards.
The job-matching platform had too many job seekers and not enough employers, and it couldn't serve women with no prior job experience or education who needed work most urgently. Data labeling emerged as a skill anyone could learn, even with basic connectivity and experience, while providing commercially viable value to AI companies building real products.
Workers in vulnerable communities may be paid extremely low wages, work in poor conditions, and experience mental health impacts from repetitive content (including disturbing imagery) - risks that remain invisible to end users because they sit behind the final AI application, similar to exploitation in fashion manufacturing.
Understand that tools like ChatGPT continue learning from every user query you submit, making you part of the training pipeline; seek transparency about where labeling work happens; and engage vendors who pay fair wages and protect worker welfare rather than racing to the lowest cost.
Our reviewer’s read on each dimension, with quotes from the episode.
There are a handful of genuine insights - data annotation as an accessible entry point to the AI labour market for disadvantaged workers, the supply/demand imbalance that killed the job-matching platform - but these are surrounded by lengthy personal anecdotes, filler affirmations, and surface-level AI commentary that a regular news reader would already know. Useful content per minute is low.
I was just trying to figure out what are the jobs in tech that anyone can learn to do, even with the most basic connectivity, the most basic skills, the most basic experience
we were so overrun with supply. So many job seekers needed jobs and we just couldn't reach enough entrepreneurs to give them jobs
The framing of data annotation as a social-impact on-ramp for women in the Global South is a modestly interesting angle, but the AI-as-social-media-gold-rush comparison, the women-in-tech empowerment narrative, and the Western-wages-vs-local-wages moral dilemma are all well-worn territory with no genuinely contrarian or first-principles argument added.
it is a bit like that like i mean... it is like a paradigm shift right it's something that is changing the direction of how we do everything
it's very much like manufacturing or fashion industry, right? A lot of big brands have been called out because of the conditions
Gina Romero is a genuine practitioner who built and pivoted a real organisation, raised angel capital, and partnered with UN Women - she has done the thing rather than just theorised about it. However, scale remains modest and unquantified in outcomes, limiting her caliber for a hard-nosed B2B operator audience.
we actually raised some angel funds, about half a million dollars. And we built the platform. We had it in beta and it was working
we actually launched together with UN Women
A few concrete data points appear - half a million dollars in angel funding, Metro Manila minimum wage as the pay benchmark, UN Women as a named partner, and the pandemic timing of the beta launch - but outcomes, user numbers, revenue, and client names are entirely absent, leaving most claims unsubstantiated.
we actually raised some angel funds, about half a million dollars
We take the Metro Manila minimum wage and we pay them that regardless of where they are in the country
The host repeatedly centres himself - referencing his own employees, his ChatGPT habits, his dyslexia, and his personal history with the guest - and offers no meaningful pushback on any claim. Questions are leading or rhetorical, and the one genuinely interesting thread (the dark side of data annotation labour conditions) is dropped without any probing follow-up.
that was a big meaty point
i used to chat gtp to you know what should i put in my password today
Computed from the transcript - who did the talking, and the words that came up most.
Connected Women, led by Gina, is making significant strides in empowering women to embrace and excel in the field of technology. Starting with a focus on executive women and entrepreneurs, their mission expanded to include marginalized groups in the Philippines, offering them opportunities for growth and leveraging technology to improve their lives. A key aspect of their work is the intersection of technology and AI. The rapid advancement of AI technology presents immense potential, but it is essential to address challenges and ensure inclusivity. Gina's dedication to diversity and inclusivity is transforming the technology landscape, creating a more equitable future. With Gina's leadership and the ethos of Connected Women, the transformative power of women in technology is being recognized, paving the way for a more inclusive and diverse tech landscape. Where to find them: LinkedIn: Website: Twitter: Join 1000’s London-centred founders, creators, freelancers and makers in the #techlondon slack channel at techlondon.io
Transcribed and scored by The B2B Podcast Index.
Welcome to Tech London, a show featuring interviews with London's top creative entrepreneurs, startups, investors, design agencies, internet marketers and freelancers that make up the Tech London online community, which mostly lives on the Slack instant messaging platform. We rotate through both hosts and guests for these interviews, so you have the chance to hear from multiple perspectives on London's tech scene. hello ladies and gentlemen and welcome to this week's edition of the tech london podcast in the studio and look before we get into it i need to say that um the the kind of organizing idea for this podcast is that it's that people have to be in london and be london centric but today's guest is london centric she just happened to move um and she's one of the people that i know on the planet in real life that knows most about this topic we're going to approach so genie romero what are you known for what would you like to be known for oh that's a really great kickoff question um so the first question i'm known for community building um but really i want to be known for being a geek that solves wicked problems but wicked as in evil evil you know slayers or like wicked as in like wicked wicked like big unsolvable unsolvable problems things that people saying can't be solved in our lifetime and probably can't that's that's that's true actually and i'm forever grateful to you and bobby because um about 13 years ago when i was struggling with microsoft 365 on my mobile phone um you said have you heard of google apps and and that transform which is now known as Google Workplace, but that has transformed my existence.
So can you say a little bit about the Connected Women origin story? Because that is huge. It's a really long origin story, so I'll make the short version. So, well, Connected Women, our mission is really to empower women through tech.
So how do we get more women participating in tech? How do we get more women using tech? How do we get more women building tech? So that was kind of the beginning idea around it because I ended up in technology because of Bobby and I fell in love with technology because it just allowed me to do so much, you know, so much more than I could have done without it.
And so that was, gosh, I don't even want to count how many years back, but at least a decade and a half. And... The problem that we set out to solve was really that, getting women more involved in tech. And not women in tech as such, but more just regular people that need to use it.
So I remember the time when this kicked off, it was social media mania, right? Everyone was like, oh my gosh, we need to be on social media. And nobody really knew how to do it or how to do it properly. So it started with simple things like that, how to set up your website, how to have your social media presence done properly.
And it just kind of evolved. And when I moved to Singapore, that's when it was formally launched. And it was really focusing on executive women, businesswomen, women in startup. But having been brought up here in the UK, but I come from the Philippines.
My mom's Filipino. It suddenly occurred to me that I was working with a lot of really cool people in Singapore, doing some really interesting work with tech companies. But there are a lot of people in the Philippines that would really benefit from leveraging technology to improve their lives and not just to to get better work-life balance but to even alleviate their poverty situation for example or their their struggle to earn so we decided to come back to the philippines and and focus on those coming from more disadvantaged groups i always describe it as um and correct me if i'm wrong here is that you go back to the philippines and thought how come all these women are still flying to san francisco and london to work in hotels and they could just be here working virtually and be with their families is an anchor representation of your absolutely yeah that was the origin origin story right because my mom was a domestic worker so she was actually a domestic worker in the uk back in the 70s um and that's why i was brought up in the uk and my mom's community of filipinos in the uk at that time was was you know a lot of women and men who had left their families and their kids behind so that they could earn and send money home because there weren't really opportunities like that here in the philippines and you know it still happens today right it does and then in i've got to get this in because in in 2019 when i actually like got an idea of um what you were doing in the philippines i we hired um venice and zara who are still with us today and zara produces this podcast she sort of wandered in to do some writing and social media and it's our podcast producer and um and one of the things was i've always i've just known about outsourcing for years and like last week we had um noel who works mainly with uh programmers from the mid from eastern europe and he based in london and i just known about it but i always had this fear of um accidentally hiring someone who you know not working in uh good circumstances and is being ripped off so you know the trust of knowing the kind of operation you'll run was a massive deal there um so my next question is um how did connectic women um what's the word pivot if you like over the over lockdown because that got you to providing the service you now do yeah so this is actually we pivoted twice so this is actually version three of connected women so the first version was partnering with companies like google microsoft um out of singapore at that point other tech companies and collaborating to get women using their tours, right?
So that was the first sort of like iteration of connected women. But like I said, most of the women were executive women, women entrepreneurs coming from a relatively privileged background. So I came back to the Philippines because my first idea was there's so much talent in the Philippines. There are so many entrepreneurs globally who need talent.
I can match, you know, women from the Philippines who want to work from home with entrepreneurs all over the world. So the first model was a job matching platform. I still think it's a really cool idea for the record. It's really hard to build a job matching platform that's focused on a very niche group like this.
And we built it. We actually raised some angel funds, about half a million dollars. And we built the platform. We had it in beta and it was working.
So when the pandemic hit, we had just launched the beta. But when I say it was working, it was working in a way So we had two major challenges that we recognized that were happening when we launched the beta. The first challenge was that we were so overrun with supply. So many job seekers needed jobs and we just couldn't reach enough entrepreneurs to give them jobs.
And the second thing that we noticed is while we have immensely talented, experienced people in our community, we also had a lot of women who didn't have opportunities to get educated, didn't have opportunities to get experience in jobs and were actually couldn't compete. They were at a real disadvantage. And these were the ones that really need the money. These were the ones who couldn't put food on the table.
So I had a bit of a moral dilemma as a social entrepreneur who set out to solve a problem where I realized if we carried on on the same route that we wouldn't solve the problem that we set out to solve. And at the time I was obsessing over companies who were doing something called data annotation. So they were working with organizations from disadvantaged areas like Kenya or Africa or India or wherever else, and even people from slum areas, and training them in something called data annotation, which allowed them to get a job in the growing, really, really fast-growing artificial intelligence industry so that piqued my interest and what is the um because this is one of the things i wanted to speak to you about is um i you know i used to chat gtp to you know what should i put in my password today to you know loads of how should i redesign the future of my business and stuff like that but um what's it's only recently i've worked i've as i've looked more into it is i realized how chat gp gets the information in there so it's it's this annotation but what does that entail and what are the pitfalls because i feel i feel like this is a you know the first world is really benefiting from the you know global south if you like just feeding it information and we just take it for granted and think it's amazing technology and all star trekky but there's there's a bit of a dark side to it and what do we need to look out for as we get excited about chat gtp and things like it yeah so it's a really interesting area um i didn't know anything about ai although i've always been a bit of a geek i like my tools and my platforms but um i really was interested because when i was looking for some you know like skills that could be learned by anyone that was really what I was looking for.
I wasn't looking to get into AI. I wasn't looking to get into any particular field. I was just trying to figure out what are the jobs in tech that anyone can learn to do, even with the most basic connectivity, the most basic skills, the most basic experience, like something that anyone can learn to do so we wouldn't have to turn people away who had passion and talent, right? Or potential at least.
And so I came across companies that I've pioneered in this space building out kind of, you know, like BPOs, like outsourcing centers or call centers that specialize in training AI tools. I've got my mind a little bit blown there. So how does this affect the - I'm trying to think of an elegant way to say how does this affect the job market, which sounds like a really obvious question, but how does that affect I think in my solution in the you know Western society and the rest of the world Yeah So basically I wanted to find a skill a technology skill that is future ready that anyone could learn And so that's how I kind of came across the data labeling industry, because there were some pioneers in this space who were doing really, really good work in providing needed services, like commercially viable needed services to companies that are building AI tools, but were also able to provide jobs to people from disadvantaged groups.
And so that was sort of the model that I looked at. And what a lot of people don't realize about AI is to train any kind of AI, any kind of artificial intelligence, you really need to just feed the program with millions and millions and millions of data or data sets. So for example, to train a self-driving car, you have to feed those self-driving cars with everything that a human would see on the road. It's called computer vision.
So teaching a computer how to see like a human. And you can imagine how complex that is because people, objects, parts of the road, traffic lights, whatever you might see through your human eyes looks different from different angles. so data labeling is basically taking millions of images and tagging it with text putting a box around it and identifying what those objects are and feeding it into machines so they can learn and you can do the same with text so text-based ai tools like chat gpt at one point were trained on you know data sets and the data was tagged to contextualize it so they can learn um a lot of models now can actually learn from users so when we're using tools like chat gpt now we're essentially still training those models and it continues to learn yeah that was um that there's a bit in um full fact by these i think he's an austrian or swedish guy that did all the tech talks hands rosling i think his name is and he points out that you know the difference between i'm gonna mess it up here but like you know for me to double my pay would be like amazing and for someone else doubling their pay in i mean not maybe not in the philippines but um or maybe it is in the philippines you know is it get getting a dollar more a day is the difference between their child going to school or not and it's it's you know you have to i'm really glad you said that because you do you do have to put it in in perspective um so just just is there anything else you want to add to what you just said because that was a that was a big meaty point yeah so well there's a lot of different skills and talent that are needed obviously to build out sophisticated tools like ai um so on one hand on one end of the spectrum you'll have your your data scientists your ai experts your model trainers um your machine learning experts your your um computer vision experts and all the rest and your programmers um and then on the other hand you need you just need to feed and tag um the the software or the applications with data right um so it makes sense to like like any kind of um you know organization or any kind of process you're always looking to outsource the cheapest labor for each part of the job, right?
And because the annotation component can be done en masse and cheaply when you outsource to certain parts of the world, right, where the cost of living is just cheaper, it means that you can actually really reduce your costs when you're doing something that requires such huge scale. The challenge with that is a lot of the time the people that participate in this market are disadvantaged, right, or vulnerable. So they will work for next to nothing in conditions that are not necessarily ideal.
So there have been some incidents where inadvertently, right, or unwittingly, organizations have set up these types of call centers or data labeling centers where the workers themselves were not treated very well because they sit so far behind the applications that we interact with, you wouldn't actually know that as a user. So if you think about it, it's very much like manufacturing or fashion industry, right? A lot of big brands have been called out because of the conditions that they have let their workers work under.
But it's a really difficult topic, Bernie, honestly, it's such a difficult topic to to talk about because we also need to be careful not to judge by Western standards, right? Because $2 a day or $3 a day might sound like pittance to a London-based company and way below what we would expect to pay anyone. But then again, some of these people are coming from zero. So it's a really fine, tricky moral dilemma to get it right.
and um what is like as you've been a you know immersed in this like i it just it just seems really this what happening right now reminds me of the days when we first met when twitter come out and instagram didn even exist and and there was this like social media week gold rush in london and everyone was going oh do you need a facebook page i'm going to cancel websites and just have a facebook page that's and there's rather bad advice like that and the only thing that survived is email but um it feels it feels like that gold rush and the gold rush of the sharing economy and but but even faster so like just what's your feeling about the the pace of what is happening at the moment yeah well i think yeah it was a bit it's a bit negative isn't it i want to lighten it up a little bit so the way that we work here at connected women is we want to give women jobs that they can do from home um so that's a big part of what we do empowering women with flexible home-based work that they can do in between the other things that they need to do.
And we actually launched together with UN Women. So we're very well vetted from a women's empowerment perspective. And in terms of making sure that people are well looked after, it's pretty basic, really. We just make sure that we pay them a decent wage.
So Philippines minimum wage. We take the Metro Manila minimum wage and we pay them that regardless of where they are in the country. And then we make sure that the type of projects and the type of work that we take on are not going to be detrimental to their mental health or their wellness. And then we teach them how to practice good practices like taking, you know, working decent amount of hours and taking breaks and all the rest.
So it's not that complicated when you break it down. It is really exciting. I mean, there are negatives to it and everything, but um i think one thing i learned from that um because around that basically like 10 years ago there was like social there was podcasting social media and sharing economy and co-working so there's like four i don't know i like the words in my little universe there were four definite paradigm shits it was a real way to like find voice and you know that technology and i'm really dyslexic So all these tools coming along, even Trello just like helped me just get a grip on, you know, how to interact with the world.
And it was life changing. And I think that set up for what I'm doing now or how, not what I'm doing now, so it makes me sound like I know what I'm doing, but okay. You know, appreciating the way that is going on with AI and what to look for. So I'm really, really grateful for that.
what is um really exciting because although a lot of ai experts out there knew that this was going to be huge i don't think anyone expected it to to go mainstream so quickly right i mean last year um or for the last few years meta and zuckerberg have been talking about the metaverse and it was supposed to be the next big thing right and then all of a sudden ai came out of the blue with chat gpt um being launched by open ai and it just went mainstream so quickly like everyone is using it even here in the philippines like chat gpt speaks such great tagalog because so many philippines are using it and pumping it with data um but i think it's really really interesting and i'm glad that you went back to that time long ago when we when we met because it is a bit like that like i mean i don't want to i don't really want to use um big words that i don't understand but um it is like a paradigm shift right it's something that is changing the direction of of how we do everything and so it's exciting but it's also scary for a lot of people i think so gina thank you very much for your time today where can we find you online and what's the best place to look for you and anything you want to draw our attention to um so you can find me on linkedin um not really on twitter anymore much um but feel free to reach out connect on linkedin we share a lot of information about the impact that we do so if you're interested in social impact and inclusive innovation um then do follow me ping me a message if you want to chat and if you want to find out more about connected women you can check us out at connectedwomen.
com thank you so much for having me bernie we'll put links in the show next to all of that and thank you gina because like i've always loved your zest and vigor for life and business and you know you're one of the people that has just genuinely made an impact in a very cool way in the world so thank you very much and you can go and blush now um ladies and gentlemen thank you for your earbuds today and whatever you're doing listening to us if you are go to the tech london.io there is a slack channel full of creators and startups and independent economic agents.
It's been going since 2014 when Jonathan accidentally started a Slack channel and has grown into thousands and thousands of people in a London-centric help out people on their path to building something great. Be careful out there. It is a jungle. over at techlondon.
io. Till next time.
Other episodes covering the same guests and topics, from across The B2B Podcast Index.