Live data from Hacker News

EU Artificial Intelligence Act

artificialintelligenceact.eu

11–20 of 129 posts

Re: EU Artificial Intelligence Act

#11
Article 10 requires that

> all training data be "relevant, representative, free of errors and complete."

This is especially interesting to me with regard to something like ChatGPT. As we know, ChatGPT occasionally gives factually incorrect information. Does this mean that, in its current form, it would be illegal in EU? We know that Google is currently blocking access to Bard in EU. Will ChatGPT be forced to follow suit?

ChatGPT is great and I love it. It would be a shame if I'm not even allowed to use it _at my own risk_ just because it might be wrong about some things. This seems like a simplification, but it sounds like EU is allowing Perfect to be the enemy of Good.

Re: EU Artificial Intelligence Act

#12
post #6

The Annexes ( https://artificialintelligenceact.eu/annexes/ ) contain a definition of "High risks AI systems" at Annex III. -- Incidentally, for the many who claimed on these pages that we would not "have a definition of AI" (actually we have several), well, this legislative text provides one: software with the ability, for a given set of human-defined objectives, to generate outputs such as content, predictions, rec…

Is there any recommendation engine which would not count as AI according to this definition?

The HN definition torture engine would agree, depending of course on which knobs you turn.

Re: EU Artificial Intelligence Act

#13
post #6

The Annexes ( https://artificialintelligenceact.eu/annexes/ ) contain a definition of "High risks AI systems" at Annex III. -- Incidentally, for the many who claimed on these pages that we would not "have a definition of AI" (actually we have several), well, this legislative text provides one: software with the ability, for a given set of human-defined objectives, to generate outputs such as content, predictions, rec…

“Symbolic reasoning” used to make “decisions which influence the environment with which the system interacts” could describe almost all real world computer systems, could it not?

Re: EU Artificial Intelligence Act

#14
post #6

The Annexes ( https://artificialintelligenceact.eu/annexes/ ) contain a definition of "High risks AI systems" at Annex III. -- Incidentally, for the many who claimed on these pages that we would not "have a definition of AI" (actually we have several), well, this legislative text provides one: software with the ability, for a given set of human-defined objectives, to generate outputs such as content, predictions, rec…

[dead]

Re: EU Artificial Intelligence Act

#15

There was a great article on this recently that cuts through the EU’s dressing: EU AI Act To Target US Open Source Software https://technomancers.ai/eu-ai-act-to-target-us-open-source-... TLDR it imposes ridiculous constraints on github and open source developers

That site has some questionalble views, for example https://technomancers.ai/pardon-elizabeth-holmes/

Re: EU Artificial Intelligence Act

#16
post #13
post #6

The Annexes ( https://artificialintelligenceact.eu/annexes/ ) contain a definition of "High risks AI systems" at Annex III. -- Incidentally, for the many who claimed on these pages that we would not "have a definition of AI" (actually we have several), well, this legislative text provides one: software with the ability, for a given set of human-defined objectives, to generate outputs such as content, predictions, rec…

“Symbolic reasoning” used to make “decisions which influence the environment with which the system interacts” could describe almost all real world computer systems, could it not?

Yup, it’s so laughably broad that almost any software that a user interacts with could come under that definition.

Re: EU Artificial Intelligence Act

#17

Why is everyone in a hurry to regulate AI? I can't think of one example where someone was harmed by an LLM. Besides "AI" is largely a marketing term, most software has "AI" elements and that has been the case for a while now, this thing has "unintended consequences" written all over it.

They're regulators. That's what they do. That's all they do.

Re: EU Artificial Intelligence Act

#18
post #15

There was a great article on this recently that cuts through the EU’s dressing: EU AI Act To Target US Open Source Software https://technomancers.ai/eu-ai-act-to-target-us-open-source-... TLDR it imposes ridiculous constraints on github and open source developers

That site has some questionalble views, for example https://technomancers.ai/pardon-elizabeth-holmes/

Well the draft proposal it references seems official, I haven't skimmed through it yet to see if it checks out or not: https://www.europarl.europa.eu/meetdocs/2014_2019/plmrep/COM...

Re: EU Artificial Intelligence Act

#19
post #6

The Annexes ( https://artificialintelligenceact.eu/annexes/ ) contain a definition of "High risks AI systems" at Annex III. -- Incidentally, for the many who claimed on these pages that we would not "have a definition of AI" (actually we have several), well, this legislative text provides one: software with the ability, for a given set of human-defined objectives, to generate outputs such as content, predictions, rec…

Is there any recommendation engine which would not count as AI according to this definition?

> Is there

Yes (oddly enough): arithmetic based.

Re: EU Artificial Intelligence Act

#20
The act is dated 21. 4. 2021. More than two years old.

1.) being a part of the team working on this has to be among the most exciting legal jobs in Brussels

2.) I did not have time to read the entire act, not even sure if I'd understand it, but I'd be curious how much of it is still relevant given the leaps in both tech and especially popularity in the last two years.

Post reply on HN