pull down to refresh

30 sats \ 2 replies \ @optimism 23 Jun \ parent \ on: Swedish PM Calls For a Pause of the EU’s AI Rules AI
The biggest problematic overarching goal - not rule - is "environmentally friendly". It's the good old German Green Agenda that puts Greenpeace dictating optimization before there even is an economically viable solution. Last I understood this measures in KWh of course, because evil network operators could be like the Germans too and use coal. Countries like Sweden that have a renewed interest in as-clean-as-possible nuclear energy (and plan to generate a lot of excess) see this as threatening, and I personally would say that that's for a good reason. After all, EU wants to ban bitcoin mining too, because energy in Germany after their self-harm is so damn expensive.
Also of note (emphasis mine):
AI systems that negatively affect safety or fundamental rights will be considered high risk and will be divided into two categories:
  1. AI systems that are used in products falling under the EU’s product safety legislation. This includes toys, aviation, cars, medical devices and lifts.
No one will argue against protecting children, but the latter 3 may also hurt. It's good to be skeptical about the EU naming something specifically because it often doesn't end well (as in: you get a no for anything) and this causes the feared "falling behind": "sorry we shall not cure your cancer because the required AI is in a 75 year testing procedure due to safety concerns"
  1. AI systems falling into specific areas that will have to be registered in an EU database:
Management and operation of critical infrastructure
wait wat?
Education and vocational training
wait wat?
Employment, worker management and access to self-employment
wait wat?
Access to and enjoyment of essential private services and public services and benefits
WITAF?!?!
Law enforcement Migration, asylum and border control management Assistance in legal interpretation and application of the law.
write
preview
100 sats \ 1 reply \ @SimpleStacker 23 Jun
Dang...
You know, I can see how this happens. I sit on an AI strategy committee at my university, and there are certain individuals on the committee that bring up all these exact points. But the problem is they never really have any specific ask, they're just like, "Well there are all these risks here..." It basically just amounts to complaining and virtue signaling. "Look at me, I'm so thoughtful"
I really just wish these people would just learn to think properly. Yes, we know there are some risks, but unless you can articulate what the risk is and quantify it to some extent, how do you expect us to take any action on it? So let's focus on what we can take action on instead of wasting our time talking about ephemeral things!
reply
80 sats \ 0 replies \ @optimism 23 Jun
My all-time favorite risk manager had the motto "never let a good crisis go to waste". Opportunistic but not to extract the monetary gainz, but to extract wisdom. The problem in the EU is similar to your committee: these are all functionaries (or worse: populists) that have no clue at all what they are talking about nor do they have experience in risk management, so because of their fear they pre-empt "rules". There's nothing pragmatic about it.
I don't think that it's malice though, it's just good old incompetence mixed with power hierarchies. It's almost unthinkable right now in Europe to leave something up for regulation at the member state level to see what works and only then look for harmonization. Everything has to be "federal" and this causes a lot of grief. When I left Europe - almost 2 decades ago now - the EU was about collaboration. Nowadays, I shake my head, because it seems to be mostly about dictation.
reply