Unknown biases

I asked ChatGPT 4o to imagine it was a critical journalist writing an op-ed article about Ai's adoption in our daily lives. Here's what it answered.

Artificial intelligence, celebrated for its convenience, poses risks to our privacy and independence. It often functions opaquely, increasing biases and surveillance. Without careful oversight, adopting AI could mean losing human control and ethical standards in favor of machine-driven simplicity. This requires thorough scrutiny.

There are three levels of artificial intelligence, and we default to the third one in our daily lives: using technology developed by others for us.

Thanks to the arms race approach to research and innovations that big tech companies and academic institutions are undergoing, we have access to the key innovations that enable these algorithms to learn from our data and generate such sound output.

But even when the code and datasets are open, it will feel like we are interacting with a black box with unknown biases.

We don't have the time or knowledge to understand those widely available algorithms.

So we'll be left with whatever companies and institutions bring to our hands.