Unaccountable, unfair, poorly managed deployments of such everyday AI systems have the potential to cause enormous harm.
We’ve seen Robodebt harm hundreds of thousands of Australians. Sexist AI-enabled job recruitment tools, AI-enabled stadium security systems that let a gun into the Melbourne Cricket Ground, racist policing algorithms, and AI meal-planners that recommend poisonous recipes are just a few among thousands of examples of everyday AI harms and risks.
Yes, we are focused on the right risks.
The article seems to believe not, but I can’t seem to find any strong justification for their argument.
The prime minister’s announcement focuses on “AI safety”, “AI security” and “frontier” AI models and companies. But Australians may remain unprotected from known AI harms that – though less spectacular – are already widespread and serious.
Nearly half of the recommendations are about increasing Australia’s preparedness for AI-related cyber incidents.
The recommendations make sense. The review is right to emphasise the role of building capacity, relationships, and cooperation. Effective and responsive AI regulation must go beyond the government simply issuing and enforcing rigid rules.
Despite these strengths, the government seems to be taking a narrow view of what “AI safety” means.
AI and algorithmic systems are increasingly integrated into important services such as healthcare, insurance, housing, finance, employment, education and social security.
We’ve seen Robodebt harm hundreds of thousands of Australians. Sexist AI-enabled job recruitment tools, AI-enabled stadium security systems that let a gun into the Melbourne Cricket Ground, racist policing algorithms, and AI meal-planners that recommend poisonous recipes are just a few among thousands of examples of everyday AI harms and risks.
As the Medicare incident showed, risks also emerge from interactions between software systems, organisations and real world contexts. It doesn’t make sense to just focus on powerful AI models out of context.
Seems pretty clear to me. Stuff like robodebt killed people are severely injured who knows how many. Nothing about systems like that, or say facial rec injuring privacy and falsy accusing people is being covered.
Instead it’s about the stupid nerd rapture critihype stuff.
The average person is more likely to be hurt by automated resume sorting discarding theirs because of their name or variations in how they write due to racism encoded in the training data than some “rogue” chatbot pwning email.
I think you’re mixing up automated decision making with AI.
Everything these days is branded as ‘AI’ as a marketing gimmick, even if there’s no real intelligence behind it.
So if you want human oversight over everything, everywhere, all the time, then no, the latest guidelines don’t help in that regard.
But to dismiss all addressed concerns as
stupid nerd rapture critihype stuff.
Is to be willfully ignorant, and not constructive.
It’s not me, you’re arguing with someone with a PhD in the area lol. The conversation only accepts articles from faculty or active PhD students.
You’re also strawmanning the author’s position, but if you’re one of the P(doom) dorks you should be all about steelmanning.
There is no intelligence behind language models. They don’t even have statefulness.
This is the internet,I don’t care who wrote it, I only judge arguments by their soundness.
And even then, my issue isn’t with their arguments, which I haven’t addressed (I don’t know why you think I strawman-ed them), it’s that they don’t have one.
At best, they’re saying that the latest AI guidelines don’t cover risks the government previously identified and already took action on, pending review.
How does that contribute to the narrative that the new guidelines are incorrectly focused?
It smells of the Chewbacca’s defense, unless you can enlighten me with a better reading.
I’m no luddite, I have a grad dip in data science and had such high hopes for machine learning. But all I can say these days is, Fuck AI and the 5 legged horse it rode in on. It’s a cancer that needs to be excised before it’s too late.
