The main lesson of thirty-five years of AI research is that the hard problems are easy and the easy problems are hard. The mental abilities of a four-year-old that we take for granted – recognizing a face, lifting a pencil, walking across a room, answering a question – in fact solve some of the hardest engineering problems ever conceived… As the new generation of intelligent devices appears, it will be the stock analysts and petrochemical engineers and parole board members who are in danger of being replaced by machines. The gardeners, receptionists, and cooks are secure in their jobs for decades to come.
What was the ratio of phone time spent talking to human vs computer receptionists when Pinker published this quote in 2007? For that matter, how much non-phone time was being spent using a website to perform a transaction that would have previously required interaction with a human receptionist?
Pinker understood AI correctly (it’s still way too hard to handle arbitrary interactions with customers), yet he failed to predict the present, much less the future, because he misunderstood the economics. Most interactions with customers are very non-arbitrary. If 10% need human intervention, then you put a human in the loop after the other 90% have been taken care of by much-cheaper software.
If you were to say “a machine can’t do everything a horse can do”, you’d be right, even today, but that isn’t a refutation of the effect of automation on the economic prospects of equine labor.
The gardeners, receptionists, and cooks are secure in their jobs for decades to come.
Except that in exponentially-increasing computation-technology-driven timelines, decades are compressed into minutes after the knee of the exponential. The extra time a good cook has, isn’t long.
Let’s hope that we’re not still paying rent then, or we might find ourselves homeless.
The main lesson of thirty-five years of AI research is that the hard problems are easy and the easy problems are hard. The mental abilities of a four-year-old that we take for granted – recognizing a face, lifting a pencil, walking across a room, answering a question – in fact solve some of the hardest engineering problems ever conceived… As the new generation of intelligent devices appears, it will be the stock analysts and petrochemical engineers and parole board members who are in danger of being replaced by machines. The gardeners, receptionists, and cooks are secure in their jobs for decades to come.
Stephen Pinker, Wikipedia/Moravec’s Paradox
What was the ratio of phone time spent talking to human vs computer receptionists when Pinker published this quote in 2007? For that matter, how much non-phone time was being spent using a website to perform a transaction that would have previously required interaction with a human receptionist?
Pinker understood AI correctly (it’s still way too hard to handle arbitrary interactions with customers), yet he failed to predict the present, much less the future, because he misunderstood the economics. Most interactions with customers are very non-arbitrary. If 10% need human intervention, then you put a human in the loop after the other 90% have been taken care of by much-cheaper software.
If you were to say “a machine can’t do everything a horse can do”, you’d be right, even today, but that isn’t a refutation of the effect of automation on the economic prospects of equine labor.
Except that in exponentially-increasing computation-technology-driven timelines, decades are compressed into minutes after the knee of the exponential. The extra time a good cook has, isn’t long.
Let’s hope that we’re not still paying rent then, or we might find ourselves homeless.