WLP379 Booknotes: Code Dependent: How AI Is Changing Our Lives
Another episode from the season Booknotes, where Pilar shares the notes she made while reading a book, allowing new insights to emerge spontaneously behind the microphone.
In today's episode:
Part 1
Pilar shares her booknotes from:
Code Dependent: How AI Is Changing Our Lives
by Madhumita Murgia
“The problem that we have is not the algorithms, but the humans.”
It would be strange to have an episode about AI and not use generative AI for the show notes, something I (Pilar) haven’t been doing for a while. So here is the summary of why I’m hesitant to advocate for the use of genAI, even though I use it myself, with some arguments borrowed from the book featured today.
My hesitancy is not rooted in a lack of appreciation for the technology's utility, but rather in a deep concern for the ethical, social, and human costs associated with its development and deployment.
Key insights highlighted in Code Dependent:
• Exploitation of Vulnerable Workers: The "data workers" behind the scenes are often vulnerable individuals struggling to make ends meet, who moderate content or tag data under poor working conditions.
• Digital Colonialism: Vast amounts of data are extracted from citizens in the Global South with "maximum value extraction" by tech corporations. This often results in expensive technology that benefits those who can pay while excluding the marginalised communities whose data helped build the systems in the first place.
• Concentration of Global Power: I’m concerned that the powerful companies behind AI are almost exclusively US-based, giving a single country immense power and an "outsized impact" on the rest of the world.
• Prioritising Profit over Social Good: Even when AI has life-changing potential, such as in rural healthcare, these services are often offered as premium packages for those who can afford them, rather than being used to close gaps for the underserved.
• Lack of Human Accountability: Algorithms are often used to tackle “messy human problems” and "fix" individuals (such as predicting teenage pregnancy) rather than addressing the underlying social fractures. Furthermore, often no one takes responsibility for the outcomes or failures of these automated systems.
And don’t forget to check out Connection and Disconnection in Remote Teams, the book!
Part 2: Personal Updates
Pilar shares some of the ways in which she’s using generative AI to help with the writing of her book.
If you’re interested in All the World’s a Workplace, sign up here.