More AI Hysteria
So, after a spate of IT outages in the US, the cry goes up - as it does with monotonous regularity these days - that it's all the fault of AI, which, in certain quarters, has become the go to scapegoat for anything bad that happens. As I've noted before, the reality of just about every actual AI-related problem I've seen reported on, human error - in that the AI was simply carrying out a task set for it by humans without sufficient parameters having been set - is generally the root cause. Look, the fact is that AI is one of the most over-hyped technologies ever. In truth, it is still very limited in what it can actually do, but the companies that develop have to keep spinning their yarns about how it could be developed to do, well, everything, in order to keep that investment flowing in. As it stands, it isn't going to be killing us all any tine soon, if ever. Job losses as a result of AI implementations are far less severe than predicted because it is so limited. Its practical applications are still limited. It certainly isn't 'evil'. 'Evil' after all, implies a concept of morality, which AI can't have, because it isn't a living being, much less sentient. It is simply a bunch of algorithms. It isn't alive, it doesn't care if you are polite to it or not, it doesn't have feelings that can be hurt - so the anti-AI brigade really needs to get a grip and stop anthropomorphising what is, in essence, simply a tool.
But they won't. Their criticisms become more hysterical by the day. Just recently, for instance, I've been seeing much pearl-clutching over a soft toy for kids that incorporates a link to some AI or other, with people reporting that it will describe horrible atrocities and sexual stuff to kiddies. Oh, the horror of it all! When you dig deeper, you find that, yes, it will regurgitate information undoubtedly inappropriate for young children, but only if it is asked specifically about them! Now, while I wouldn't disagree that such devices should incorporate safeguards to protect children from exposure to inappropriate material, the fact is that the average three year old is highly unlikely to be spontaneously asking about such stuff, so the risk is largely theoretical. It's an 'outrage' largely manufactured by certain media actors looking for an anti-AI scare story. I mean, if you put can of soft drink in a microwave and heat it up, it will explode, but that doesn't mean that canned soft drinks should be banned as being potentially dangerous to children who might put them in microwave ovens. Manufacturers assume that most people just aren't going to let that happen. Look, I'm not here banging the drum for AI and its developers, but these kind of idiotic scare stories ultimately serve only to distract from the very real and present problems posed by the current models of AI, namely their consumption of resources - the power and water hungry data centres they need to operate, which blight the environment and contribute to global warming. That's the problem that needs to be addressed. We need to focus on pressuring AI developers to come up with models that don't require such a disproportionate use of resources, that are more energy efficient. It's in their own interest, after all, as pretty soon this will become a natural barrier to their AI models' further development - there simply won't be enough resources available to run them. Ao, anti-AI luddites, dial down the hysteria and focus on the immediate issues at hand!
