Well 2001 is kinda unlikely at this point, yes. On the other hand, I would be surprised if it took much beyond 2050 for us to get it working. 3001 is insanely far in the future.
Ah, I forgot to mention that I had his course in spring 2001 ;)
His evaluation was based on his experiences from some 30-40 years of doing AI work and following the progress made. And with the (lack of) progress that's been done in natural language recognition he thinks it's more likely to take a thousand years than a hundred.
What do you think will happen within the next 40-50 years that will be a huge leap compared to what's happened in the past 40-50 years?
Not saying that nothing's happened in the last 40-50 years, of course, just that it will take more than 40-50 years to get to real, actual talking computers/robots.
We've been making HUGE strides in AI lately. In just the last 5 years we've gone from driverless cars that couldn't keep to a simple dessert road for a few miles, to successfully navigating towns with intersections, traffic lights and simulated pedestrians. Most of these strides have been made, BTW, by throwing out most of what the folks of the last 30-40 years were working on...
Besides, with the rate of acceleration of technological change we should see roughly 20,000 years worth of change (at year 2000 rates) by the year 2100. We'll see 1000 years worth of change (at year 2000 rates) before 2070 or so.
Nah, that's according to some idjits. There's at least one Mayan inscription that says that King Pacal of Palenque wants the 18th Calendar Round of his accession to be celebrated. That won't be until 21 Oct 4772.
Clearly they could count higher than they usually used on their inscriptions. They just didn't usually bother.
![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
swestrup
![[identity profile]](https://www.dreamwidth.org/img/silk/identity/openid.png){kind=small}
no subject
Date: 2008-07-29 09:07 am (UTC)My old (ancient, actually) AI teacher at university told us that 3001 was far more likely than 2001 for HAL like artificial intelligence.
no subject
Date: 2008-07-29 04:34 pm (UTC)no subject
Date: 2008-07-29 08:22 pm (UTC)His evaluation was based on his experiences from some 30-40 years of doing AI work and following the progress made. And with the (lack of) progress that's been done in natural language recognition he thinks it's more likely to take a thousand years than a hundred.
What do you think will happen within the next 40-50 years that will be a huge leap compared to what's happened in the past 40-50 years?
Not saying that nothing's happened in the last 40-50 years, of course, just that it will take more than 40-50 years to get to real, actual talking computers/robots.
no subject
Date: 2008-07-29 08:33 pm (UTC)Besides, with the rate of acceleration of technological change we should see roughly 20,000 years worth of change (at year 2000 rates) by the year 2100. We'll see 1000 years worth of change (at year 2000 rates) before 2070 or so.
no subject
Date: 2008-07-30 01:49 am (UTC)no subject
Date: 2008-07-30 02:22 am (UTC)Clearly they could count higher than they usually used on their inscriptions. They just didn't usually bother.