The EU AI act in UNDER 3 minutes

The EU AI act in UNDER 3 minutes
artificialintelligenceact.eu/...

[@darthutah6649]

The EU Act actually doesn't seem like a bad idea. Hopefully it's not implemented in a manner which hinders innovation.

[@leonsvideos]

Great video! Looking forward to the next one. Subscribed! A suggestion: Could you make your voice a little louder? It is a bit quiet compared to the background music. And maybe go a little closer to the microphone to get better quality :)

[@Knowledgebrief128]

No problem man. Thanks for the pointers!

[@tomaszantecki4292]

I'm worried this could lead to AI companies offering inferior products in the EU compared to elsewhere. This might quickly widen the gap between the US and EU, affecting even unrelated businesses. They’ll struggle to improve processes with AI, falling further behind in competitiveness.

[@Knowledgebrief128]

Nice comment. Thanks man. I’m going to do a video on US AI policy and discuss some of the benefits of ‘innovation first’

[@jjfdkdjiwejnd092]

As the EU we might be late to dominate the AI market, but I'm glad we are leading when it comes to regulation. Because no company will regulate itself. When labour as value is taken away from the common man, he will have nothing to bargain with, and these rules will help.

[@Knowledgebrief128]

Human-complementary AI - I think this is where we need to go.

[@Leto2ndAtreides]

The Banned stuff looks like it'd cover things that should be done - albeit with some standards enforced. And the "danger to society" part all sounds iffy... And like it would either not be categorized as such, or people who build such things wouldn't bother telling the government that they're going to build it.

[@Knowledgebrief128]

Interesting stuff. I mostly agree. There’s no place for social scoring in a liberal democracy. I do believe that with biometric identification, it will still be used by police forces. Again throws up another interesting question. Will be covering a lot more stuff like this, so give me a subscribe if you’re interested!

[@Leto2ndAtreides]

@@Knowledgebrief128 Most things depend heavily on the details of the implementation. For example, under the law, repeat offenders are likely to get harsher punishment than first time criminals. That is clearly a decision based on a person's past contribution to society. Someone known to be a good person is also likely to get more leniency in many situations. A social credit score is an abstraction of that idea. - Alternatively, if we look at credit scores (of the financial kind), in general people are averse to certain kinds of information being used to feed those models. But at their core, the purpose of the models is to increase the chances of giving money to people who will successfully pay it back. The more accurate those models are, the fewer the defaults, and the better the interest rates that banks can offer (plus, less damage to the economy) In countries without common credit score systems, people either have less access to credit, or even responsible people get worse terms because the bank doesn't know how much risk it's taking on each loan. - As much as we malign China, it's interesting that many of our systems fundamentally overlap. And our politicians seem to have the same kinds of underlying instincts for how to rule (when not restricted by the law). The main difference is that let's say in the US, most of the stuff isn't consolidated, and everything is happening piecemeal... Like the example of punishments mentioned above. The Chinese system (if it ever fully matures), will have its own unique implications... Both negative and positive.

[@emehe69]

Americans are inventing and training AI Models but Europeans are inventing AI regulations, hahaha... What kind of future will Europe have doing shits like this? hahaha.

[@Knowledgebrief128]

Haha. you tell me man