Content Warning

In this Op-ed for Scientific American, Asmelash Teka & I discuss one of the many reasons the idea of replacing US federal workers with so-called generative AI systems should terrify us. We dive into the well defined task of automatic speech recognition (ASR), and describe why OpenAI’s Whisper, which has been integrated into ChatGPT, makes stuff up, or “hallucinates” as it’s called in the industry (bad nomenclature).

https://www.scientificamerican.com/article/replacing-federal-workers-with-chatbots-would-be-a-dystopian-nightmare/

Content Warning

@timnitGebru John Lithgow --> 20 Lessons on Tyranny

Now, more than ever, we need the wisdom of our intellects, the patriotism of our citizens, and the passion and talents of those who still believe in the American experiment. I am deeply grateful to Timothy Snyder for his 20 Lessons On Tyranny and for talents of the brilliant John Lithgow for bringing them to life.

Watch --> https://www.youtube.com/watch?v=cXR5HLodsT8

#tyranny #democracy #civilrights #humanrights #peace

Content Warning

@timnitGebru A hallucinating agent that might make "decisions" your livelihood depends on is bonkers, yet already happening.
I also fear the dehumanization of it.
Our interactions, like this text, is mediated by a machine, making it convenient but stripped from all the richness of a face to face conversation.
Even in bureaucratic interactions, cold by design, there are some authentic, exchanges.
We let ourselves be trapped within the enclosures of technology.
+ -

Content Warning

OpenAI’s (and Muskrat and others’) quest to build one-model-for-everything has resulted in less reliable systems than we had, even in the well defined task of ASR. Historically “hallucinations” weren’t problems in ASR systems! So now imagine what will happen if DODGE replaces federal workers with these tools to perform tasks that expert federal workers perform?

“There is no “one weird trick” that removes experts & creates miracle machines that can do everything that humans can do, but better.”

Content Warning

@timnitGebru

"Better" isn't their goal.

"Cheaper" than employing human beings is, or more accurately "more profitable", and things like accuracy and human suffering resulting from errors are collateral damage in the quest to maximize profit.

Just like industrialization before it, they're perfectly happy to accept mangled limbs and ruined lives as a byproduct to their profit. They have historically pit children in harm's way (and still do!) - causing ugly problems because of hallucinations closing bank accounts and giving fatal advice *isn't even on their radar*.

This isn't a new problem. There are certain technologies that are powerful, and inherently dangerous. Figurative chainsaws - useful in trained hands who understand the dangers and how to operate them safely - deadly in others. We license and regulate those technologies - or we used to. Maybe it's time to start again. Remember: consumer safety laws are written in blood.