AI Companies Make Yet Another “Vow” To Improve Their Practices
Here we go again. The major players in the AI arena – you know them: the likes of OpenAI, Google, Meta, and a few others – have strutted up to the White House, as if they were appealing to a school principal, to earnestly vow… drumroll, please… to tackle the misuse of their AI tools that generate fake rubbish. Yes, that familiar issue of ‘non-consensual deepfake porn’ and, even worse, child sexual exploitation materials. As if they haven’t been saying this for ages.
They must be joking, right?
Honestly, mate, how many times do we have to endure these tech titans swearing on their grandmother’s memory that they’re going to fix this? It’s like watching the same episode of *Coronation Street* over and over again, but instead of a pint at the pub, you get a load of jargon about “responsibility” and “voluntary measures.” Yeah, voluntary, just like that time I “voluntarily” chose not to finish the entire pizza last night. We all know how that turned out.
Promises, Promises: What They Really Stated
Right now, you’re probably thinking, “What did these guys swear this time, mate?” So here’s the scoop: they claim they’ll incorporate more watermarks or digital signatures into their AI-generated content to assist users in distinguishing the authentic from a deepfake disaster. The funny part is, they believe these watermarks will be so discreet you wouldn’t even notice… sure, like that won’t backfire.
Digital Watermarks: The Tech World’s Version of ‘Caution: Hot Coffee’
Watermarks, huh? That’s like putting a “Caution: Hot” label on your grandma’s tea in her modest bungalow. If we need watermarks to identify a fake, the damage is already done, isn’t it? By the time someone clicks on a dubious video or image, the harm’s already occurred, like stepping in dog mess. Sure, a watermark might inform you it’s fake, but that’s after you’ve already got the mess all over your shoes.
Another Non-Binding Commitment, Like We Haven’t Heard That Before
Oh, and here’s a kicker: none of this promise-making is enforceable. They’re just *voluntarily* pledging to do better. No laws, no penalties, nothing truly holding them accountable. It’s akin to a bloke promising to stop leaving his socks strewn about the house but still doing it the next day. These tech giants will do as they please, and when the pressure’s on, they’ll just put out a few press statements and say, “Well, we made a promise, didn’t we?” Sure you did, mates.
Who’s Really Monitoring Them?
So here’s the crux: Who’s going to step in and give them a smack on the wrist if (when) this all goes awry? If it’s not legally binding, the only ones keeping an eye on these blokes is… well, themselves. It’s like putting the fox in charge of the henhouse and expecting it to voluntarily avoid the chickens. Just a tad foolish, isn’t it?
ChatGPT, Bard, and Others – Are They Responsible?
Let’s not ignore the AI models themselves – ChatGPT, Bard, or whatever trendy name they come up with next week. These chatbots have been spouting some dodgy content for a while now. Sure, most of the time, they’re polite and helpful, but occasionally they’ll misbehave like a teenager discovering Spotify. Yeah, they’ll claim they’ve improved the safety measures, but we’ve all seen how quickly things can spiral if they’re not closely monitored.
“It’s The Users, Not Us!”
Now, of course, the big companies will argue, “It’s not our fault; it’s the users who misuse the technology!” Sure, fine, lads, but you’re the ones who designed these products, unleashed them on the public, and then looked the other way when things went south. It’s like handing your mate a hammer and wondering why the TV’s broken. You provided the tool; the least you could do is ensure it doesn’t cause chaos.
Should We Believe It? Probably Not
Look, I’m not saying they’re all terrible. AI can be brilliant for some applications – automation, creative tasks, sifting through massive data – you know, the sorts of things that simplify life a bit. But we must be wiser about this. Just because the tech giants promise something doesn’t mean we should be celebrating with streamers and confetti. Save the confetti for when they genuinely deliver, which, let’s be honest, they likely won’t.
Oh look, another voluntary, non-binding pledge to improve
Some of the largest AI companies in America have made a serious commitment to the White House to prevent their AI products from being misused to create non-consensual deepfake pornography and child exploitation material. But let’s be real, it’s the tech world’s version of your friend promising never to overindulge again after an embarrassing night out. We’ll believe it when we see it, lads.
