One Size

Most of the arguments that the author makes about the dangers of ubiquitous use of AI would hold true for other technologies that were fundamentally life-altering. The scale and nature of those innovations may have been different but the harms they could do were similar. Misinformation and lies have been propagating at decent clip long before AI became so accessible to everyone. We had good old-fashioned television and printed newspapers doing that quite effectively. 

"A car that accelerates instead of braking every once in a while is not ready for the road. A faucet that occasionally spits out boiling water instead of cold does not belong in your home. Working properly most of the time simply isn’t good enough for technologies that people are heavily reliant upon. And two and a half years after the launch of ChatGPT, generative AI is becoming such a technology."

The "heavily reliant" is the big question here. Is it okay to use the technology in a trust by verify mode? So if the said car is known to accelerate instead of brake on occasion but is otherwise quite the picture of perfection and way cheaper than any other car available in the market, maybe the customer might develop a workaround to make sure they are not taken by surprise when the brakes produce acceleration. It is a question of the perceived value notwithstanding the defects. There come a point when some degree of flaw even if a rather serious one comes to be acceptable. People have to decide for themselves where that point lies for different situations in which they use the AI. There is no one size fits all or one rule that fits all situations.