Yes, I just read an article that reeked of 'AI'. 🤢
It's really distracting - made me unsure if any of the sentences I read there are factual.
"AI systems can do some remarkable party tricks, but there's a huge difference between producing a plausible sentence and a good one. After the initial rush of astonishment, the stench of botshit becomes unmistakable."
The biggest use-case for these #LLMs (don’t call them #AI) is to produce endless reams of plausible-sounding, factually dubious content at low cost for the purpose overwhelming the information space in something akin to a DDoS attack. They will degrade our ability to search up factual information to the point that the internet will essentially become useless as a tool for educating or organizing.
But in practice, hasn't this already created the polarised "bubbles" we're suffering the effects of now? I.e. virtual balkanization, where the borders are ideological, rather than geographic?
And, if so, how do we encourage cross-pollination?
(Thinking out loud, here, mostly 😁)
I feel like, now, we are many (curated) tribes, talking past each other in cyberspace.
Fair point. I was thinking more of developing a collection of trusted sources that one can rely on to provide accurate accounting of events vs always having to wade through a sea of machine-generated nonsense.
But of course how one selects these people depends on one’s biases and political leanings, which in turn can lead to information bubbles and echo-chambers. Not sure how that’s totally avoidable.
One thing I’ve been saying for decades is that in an age of unlimited access to information (both reliable and unreliable) the most important skill we should be teaching our kids is how to critically evaluate information.
Add comment