Meanwhile, some of the loudest voices selling “AI safety” deserve antitrust scrutiny. When incumbent firms warn that the technology is too dangerous for ordinary competition but conveniently safe inside their own products, partnerships, and forthcoming IPOs...fear begins to look less like public service and more like market capture. Safety rules should if anything try to address measurable harm and not this cycles IPO no matter how many people cry in front of the media.
I know astra and claude make banging out the boiler plate nice enough to mallet into production but they aren't building trusted mechanisms and the meme they are about to transcend humanity based on cloning hugging face repositories and come kill me with my home security or destroy humanity because fiverr.com is being gutted right before a historic financial event should be scrutinized more than "i like x model more because i did a prompt that worked" it should be treated like financial fraud, these people are telling normal people they are going to die if they aren't funded and supported and allowed to gate keep 3rd parties from participating, how is this happening in the USA? I don't want to hear trump I want to hear how this is remotely legal for anyone to threaten the widespread death and destruction of humanity via surrogate vested participants if the alternative is not trusting them with their livelihoods and physical security?
Beyond offensive this, in my opinion, is illegal activity that should be absolutely destroyed and not by pushing dog shit legislation that trips up normal people building AI on any frontiers but these people from lording over non-technical people with literal death threats like they are working on the manhattan project when they are working on an interactive search engine.