You and AI - the politics of AI
ฝัง
- เผยแพร่เมื่อ 17 ส.ค. 2024
- Kate Crawford, Distinguished Research Professor at New York University, a Principal Researcher at Microsoft Research New York, and the co-founder and co-director the AI Now Institute, discusses the biases built into machine learning, and what that means for the social implications of AI.
The talk is the fourth event in the Royal Society’s 2018 series: You and AI. Supported by DeepMind.
For more information on the event series: ow.ly/PKug30jWEYV - วิทยาศาสตร์และเทคโนโลยี
Cars in the UK and Ireland have to go through technical and safety checks every few years to ensure they are still fit for purpose. Why not have a similar test for AI systems where they are tested against a dataset (suited to their purpose) of known quality and have to produce a particular score. If there are any changes to the code, you go through the test again.
Great idea, but i think the engineers are the ones needing it even more! " I am just an engineer" doesn't cut it anymore. We should open their heads to see if someone is in there!🤣🤣🤣
Thank you Kate, a very thought-provoking discussion. I think the line between using AI for the better and for the worse is very delicate. It might be easy for engineers to act without ‘morally imagine’ what are the implications of AI algorithms for real human beings and for the society as a whole. Your speech made me think whether this AI world is similar to the Hitler’s rule because ordinary people just become accomplices to big tech companies privacy intrusion, racial profiling, binary classification. This question is gonna stay in my mind for some time.
29:13 i think its quite a powerful way to look at it
by the time we finish watching & musing over this ..AI systems are learning from each other exponentially.. no sleep no compassion 🙏🧬✨
It is always nice to have a balanced view of technology, which Kate unfortunately didn’t quite achieve, although it is still important to listen to and understand different views in order to gain a systemic overview.
This is a very interesting part of the AI debate. It seems to me that it is going to take a looong time to sort this out. A sort of chicken and egg situation arises as the technology becomes available without restraint and the ethical and associated issues come later making it very difficult to change what is out there doing damage by it's misuse, etc.
I'm surprised that making data sets and algorithms open source wasn't mentioned.
The Identity Politics of AI...
Great talk and views.
I've got a feeling we're only hearing half the CalGang story.
U&I ? Qui bono ? Common misspelling of the Latin phrase cui bono? Accentuating the differences mistaking the explanans for an explanation? We could go deeper than this and add empathy ? Man more readily believe what she /he wants to be true
This woman would be telling a whole different story if Hillary was president.
The most frightening aspect of AI is the context in which it is being produced: the most uneducated, ignorant society we have seen in a long time. Wisdom, common sense cannot be "transmitted" to a machine particularly by engineers or any other "specialized" idiots.
AI must remain a tool in the hands of only those who understand the meaning of personal responsibility. And can spell it too please.
We are already experiencing "I am just an engineer", "I am just a teller" There is what seems to be a global allergy to any responsibility. The world is going to pots!
We have to ask: do we like it? If not: what are we gong to do about it?
Of course the worse of humanity is going to use this tool for the worse human endeavours! Is anyone sane surprised?
We are entering the age of "consciousness" hopefully and AI will be put in its place. It is only there to serve humankind. AS A TOOL. It should be severely restricted from anything else.
It can help with early breast cancer detection. Tick, and such other useful activities.
Look what mess it is already creating. Where is debate gone? common sense? justice? civil rights? does it recognize corruption? work at detecting it? or just help nefarious forces to enslave us further?
Thank you for disclosing what a folly it is to let a piece of engineered piece of machinery decide on our most important human values. In the name of convenience?
What? they are not enough of us to do the job anymore? MUST WE ABROGATE ALL OUR RESPONSABILITES, OUR HUMAN EVOLUTION AND ACCOMPLISHMENTS?
Binary thinking is stupid.
Caveat auditor. 10.5281/zenodo.4037538
Do no harm
Statisticians should be doing the job, not engineers and computer scientists
Blep~ :3
the girl speaker is hot