Award-winning journalist and data scientist Karen Hao flew from Hong Kong to warn the Notre Dame community about artificial intelligence.
Over the past decade, the generative AI industry has taken off, but according to Hao, there are many unforeseen consequences to this explosion, with hefty moral implications.
Hao has presented the information from her talk to Congress in the past, with a call for protective regulations in mind.
Valentina Kuskova, a professor in the Lucy Family Institute for Data and Science, attended the lecture.
“I think one of the most important takeaways from this talk is that lots of people do not realize the true cost of AI,” Kuskova said after the talk. “It’s not just the financial cost, but the human cost, the cost of suffering.”
There is not yet indication that a tool like Open AI’s ChatGPT is directly causing disparity in the world, but Hao’s presentation explained the connection between the advancements of the industry and something she coined as AI-colonialism.
In order to generate new and more advanced artificial intelligence, existing AI needs to be fed new data to learn from it. This data needs to be formatted for the artificial intelligence to be able to read it, a process known as data annotating.
As generative AI systems gained immense popularity in the last five years, large tech companies turned to developing nations and the global south for cheap labor, Hao said.
Based on Hao’s research, educated engineers in countries such as Venezuela and Kenya find themselves with no better option than to work for extremely low wages as they annotate data for tech giants like Microsoft and Open AI.
Hao said she would often meet people working for wages that were anywhere between $2 per hour if they were tenured to $10 dollars every couple of weeks if they were doing freelance work.
The conditions, Hao added, were also inhumane, consisting of long hours and no benefits.
“I never heard of the AI industry using a lot of laborers before to label data,” graduate student Bruce Huang said. “It is really stunning to me.”
The talk was part of a larger speaker series on artificial intelligence, hosted by the Lucy Family Institute for Data and Science at Notre Dame. The Lucy Family Institute takes inspiration from the University’s mission as it works towards being a center for data innovations geared towards the common good.
“We ask ourselves, ‘How can data AI technology act as more of a bridging interface towards delivering solutions that can be positively impactful on society?’” So that’s what we’d like to do at the Lucy Family Institute” founding director Nitesh Chawla said.
Visiting journalist lectures about dangers of artificial intelligence
Karen Hao delivers a presentation warning on impact of artificial intelligence to a crowd in the Jordan Hall of Science.
