KI-Minister sagt, dass OpenAI nach der Schießerei in British Columbia immer noch nicht genug unternimmt, und wird CEO Altman treffen
https://www.cbc.ca/news/politics/open-ai-tumbler-ridge-safety-policies-9.7109001
KI-Minister sagt, dass OpenAI nach der Schießerei in British Columbia immer noch nicht genug unternimmt, und wird CEO Altman treffen
https://www.cbc.ca/news/politics/open-ai-tumbler-ridge-safety-policies-9.7109001
3 Kommentare
Reading the article, OpenAI says they changed their policies and that the shooter’s account would have been flagged under the new ones.
The solution isn’t intrusive hypersurveillance from tech companies. As long as they reasonably act upon concerning behaviors, and as long as AI does not actively contribute and nourish (would-be) criminals and/or help them plan their wrongdoings, I would not blame them much.
And this comes from someone who really ates AI. I feel like OpenAI may be a “scapegoat“ in this situation. Because, as long as their AI did not actively nourish the shooter’s behaviour or help them plan to kill, most of the problem in the shooters‘ behaviour comes from the *personal mental health issues* of the shooter.
The last bit of the article is key here:
>Solomon has said all options are on the table when it comes to regulating OpenAI and other companies.
>
>Most MPs on Parliament Hill on Friday agreed that the government will need to introduce legislation that would require companies to flag problematic accounts to police.
>
>“We need to do something. What happened, the tragedy in Tumbler Ridge, I think is a lesson and protecting Canadians should be the priority of every government,“ Liberal MP Gurbux Saini told CBC News.
>
>Conservative ethics critic Michael Barrett said he would be in favour of some form of regulatory framework.
>
>“We’re going to need to really work together as parliamentarians to get our arms around this thing because we see that there are real risks, real consequences to people,“ he said.
>
>Green Party Leader Elizabeth May said any regulations that are brought in need to „go way beyond wagging a finger“ at tech companies.
>
>“Meeting those families was really hard and it stays with you,“ May said of travelling to Tumbler Ridge for a vigil earlier this month.
>
>“The fact that something could have been done if only the rich bastards in the AI industry had reported what they knew makes it a little bit harder,“ she said.
There appears to be enough alignment between various parties that some meaningful regulatory frameworks around these issues can be worked out. This is a promising development.
Open AI is another pied piper. Don’t try and come to an agreement or work out an arrangement. Don’t trust them or any other big AI outfit at all.