naivasha, kenya – Caroline Njau comes from a family of farmers tending fields of maize, wheat and potatoes in the hills near Nyahururu, 180 kilometers (112 miles) north of the capital Nairobi.
However, Njau chose a different path in life.
The 30-year-old currently lives in Naivasha, a picturesque town halfway between Nyahururu and Nairobi, the capital of Kenya’s flower industry. She’s sitting in her living room drinking milk tea and using an app to label data for an overseas artificial intelligence (AI) company. As she flips through images of paved roads, intersections, and sidewalks on her smartphone, carefully drawing boxes around various objects, the sun rises over a dirt road in her neighborhood. Traffic lights, cars, pedestrians, signs. The app’s designer, an American subcontractor for a Silicon Valley company, pays her $3 an hour.
Njau is a so-called annotator, and her annotations of data compile the building blocks for training artificial intelligence to recognize patterns in real life, in this case self-driving cars.
“My parents don’t fully embrace technology because they find it difficult to learn. But I’ve always loved science. Data annotation creates opportunities. You don’t need a degree to do this. No, all you need is a phone and an internet connection,” says Njau, who started annotating in 2021 while studying education.
Kenya has emerged as a hub for this online work, competing with countries such as India and the Philippines. The birth of technology startups since the late 2000s, the subsequent entry of technology outsourcing companies, and pro-business policies, skilled labor, and high-speed internet have created an economy where digital work is a livelihood. I did. Most of the young people. According to a 2021 Kenya Private Sector Alliance (KEPSA) study, at least 1.2 million Kenyans are working online, most of them informally.
But data annotators in Nairobi recently revealed a less optimistic side of the industry. An article in Time last year described the “torture” employees at a Nairobi outsourcing company experienced in labeling text extracted from the darkest corners of the internet. All to help OpenAI’s ChatGPT recognize harmful content. . According to the article, workers were paid less than $2 an hour for this work.
Rural AI
Despite all this talk, the annotation industry continues to expand far beyond the cramped office spaces of Nairobi.
In mid-January, when Kenyan President William Ruto launched a government-backed technology hub in Kitale, an agricultural city near the border with Uganda, young ICT students spent three weeks training AI for Silicon Valley companies. I explained how I made $284. He was using his Remotasks, his American website where freelancers get paid to label data.
A video clip of the Tech Hub, one of a series of facilities designed to equip learners with marketable technology skills, spread like wildfire on social media, with young Kenyans rushing to sign up for Remotasks accounts. Now create.
“Many young people are unemployed. Even computer science graduates cannot find jobs. The government is right to help young people access jobs online,” said a recent graduate from Eldoret in western Kenya. says nurse Kennedy Cheruyot (24).
He opened a Remotasks account in 2021 and has continued to work online while searching for a job at a hospital. Some of his friends have completely quit other careers to focus on digital tasks.
“Earlier, in our culture, boys were supposed to go to the farm and herd the cows. Now they are indoors to work online,” he said, overlooking Eldoret’s business district. Cheruyot said when we met at a cafe. Hardware stores and agricultural supply stores merge with bright yellow signs advertising Internet Cafes, the so-called “Cyber.”
Cheruyot’s dream is to own a ranch “like something out of a Western movie,” but for now he mostly looks for appearances online to pay for rent, food, utilities, water and transportation. spending a lot of time.
Commodity prices in Kenya have been soaring since 2022 due to that year’s prolonged drought and the Russian-Ukrainian war. Meanwhile, the Kenyan shilling continues to weaken due to dollar demand from energy and manufacturing industries. When the shilling falls, import prices rise, and with it the prices of consumer goods like Cheruyot.
Once he gets a job as a nurse, he expects to continue working online in his spare time, earning anywhere from $5 to $20 an hour, depending on the job.
“We don’t care if Western AI companies get rich because of our work, as long as we get paid. It may not seem like much, but it helps a lot in Kenya. “There is,” he says.
new generation of scientists
But for Njau, the tedious tasks online are the gateway to something bigger.
“Currently, commentators in Kenya are watering other people’s gardens. The flowers are starting to bloom, but we are not even there to see them,” he said outside his brick house. she said, pointing to the green grass.
“I don’t want to stop at data annotation. My goal is to advance technology. I want to know where the data goes and how the AI is programmed. We Kenyans should become data scientists,” said the CEO, who is already training people with disabilities and young women in data annotation in collaboration with Next Step Foundation, a Nairobi-based non-profit organization. says Njau. Recently, she was awarded her AI and Data Science Scholarship by the Department of Investment, Trade and Industry.
Such programs are aimed at making Kenya a frontrunner in the technological revolution, explains Next Step Foundation’s training manager, Mr. Nixon Otieno.
“I wouldn’t be surprised if Kenyans come up with the next big AI invention. We have an innovative generation and there are many problems to solve. How can we use AI to inform the Kenya Power Lighting Company about power outages?” asks Otieno.
Still, it will be difficult for Kenya and other African countries to stand out as centers of AI innovation. According to South African AI scholar and United Nations University President Professor Chilij Marwala, the education system needs an overhaul.
“Africans often receive a highly specialized education, which is the case in countries with British-oriented education systems such as Kenya and South Africa. But specialized education is outdated in an interdisciplinary world,” he says. argues, and gives an example, that creating an AI platform that analyzes X-ray images requires mastering both medicine and computer science.
Much of the conversation around AI, such as OpenAI’s ChatGPT, focuses on human jobs that carry the risk of redundancy, which is also a serious concern in African countries. But Marwala believes many people are “overstating the importance of AI and confusing it with regular automation.” Additionally, AI can help small businesses grow.
“Flower production could become cheaper for South African flower farmers if they used AI to analyze soil quality using cameras instead of paying scientists. AI “We expect this to significantly improve efficiency and reduce costs,” he says.
AI apps that rely on data labeled by Kenyans, such as the chatbot ChatGPT, are already popular among young people like Njau and Cheruyot. He finds it “really useful” when he needs recipes or travel plans. But that cannot replace his work.
