4 ms·
The whole AI space seems to suffer from this. Remember in the late 90s, big tech companies (particularly Microsoft and IBM) confidently declaring that conversin
by rsynnott 3y ago
The whole AI space seems to suffer from this. Remember in the late 90s, big tech companies (particularly Microsoft and IBM) confidently declaring that conversing naturally with a computer via voice recognition would be the primary way people use computers within years? Meanwhile, 25 years later, voice recognition is finally getting into the realm of being good enough for unimportant things where accuracy doesn’t matter.
The only mainstream AI-adjacent thing that I can think of which ever actually became “good enough” is OCR, and that took about 50 years.
I’m somewhat surprised that people aren’t more cautious, honestly. We’ve had _two_ “this time it’s different” moments in a little over a decade, now (self-driving cars and generative AI); while self-driving cars are at this point squarely in the “okay, maybe it’s not so different” bucket, I think people underestimate the risk that generative AI will, ultimately, also underwhelm.