So I figured I'm discuss some of the bets on the site.
» By 2029 no computer will have passed the Turing Test. Bet: $20,000
The Turing Test is designed to see if a person (Person B) can tell the difference between another human (Person A) and a Computer (or machine intelligence) in normal conversation usually through an Instant Messanger system or the like.
There are two approaches as to what kind of computer would be able to pass this, this first being an entity capable of human-level "thinking" or "intelligence", while the second is that the computer would outwardly appear to be human but consist much of something akin to canned responses.
The second would be something like giving the intelligent system a large amount of data to work with, such as giving it a vast library to scan, perhaps logs of chat, and respond in a similar fashion. The only problem is that this isn't true thinking. It really isn't different than asking a computer to store and recall facts, which is NOT true intelligence. You wouldn't ask "What is the square root of 15,545,492?" and get "3,942.77719" as an immediate response. A human would more likely respond with "let me get a caclulator... [time passes]... about 3,943". The problem with this, is that the program would be much less self-sufficient/self-functional and more programmed, which could pass itself off for human, but wouldn't be the depth that humans desire for the test to be beaten.
The first is much harder, and as the challenger argues, something that isn't likely to happen in the next 20 years. Much of the research and development in this territory is going towards replicating or "reverse-engineering" the human brain where current technology comes up greatly short. The entity would have to create a false life with understanding of current knowledge (which would be easy, it is just like storing and recalling facts). The problem comes from emotion, both understanding and displaying it. Since just programming a response (I.E. someone says my friend died recently, and the response is "I'm Sorry") is more on the level of the 2nd, the entity would have to be able to respond on a deeper level (such as trying to console the person).
Personally speaking, I think you could program something to respond in a human way, and we have bots that can do this to a minuscule degree. However, In my opinion, by 2029, the Turing Test will remained truly unbeaten by a computer or machine entity. I will be interested in finding out about that actually manages this in the future.
For more info on the technological limits on computing read about Moore's Law.
» By 2030, Commercial passengers will routinely fly in pilotless planes. Bet: $2,000
Pretty straightforward, no complicated explanation needed.
In my opinion, I feel that this bet will be won by the challenger. While some of the people agreeing with the statement cite the Unmanned Aerial Vehicles or "UAVs " that are in use by the military as proof that we are close to such an occurrence. The problem is that they are still piloted by humans (albeit remotely), the machines don't think, act, or even fly by themselves. They are in a certain location because a human controlled it, they are performing a certain action because a human made it.
I've always stood by a belief that no matter what, humans can't be completely replaced by machines, at some point a human will have to monitor or complete maintenance on the machines just because of our general distrust of machines doing everything by themselves without monitored by a human able to pull the plug on it. Plus, in yet another couple things that one cannot program into a machine is decision making skills and experience. In the time of an emergency (say the plane that crashed into the Hudson Bay) the pilot (and official badass) Chesley Sullenberger had 34 years of flight experience, something you just can't teach a machine, it must be programmed. Then the problem becomes when the computer can't resolve a problem between multiple rules/regulations or if the problem doesn't fall under the jurisdiction of any specific rule/regulation, humans lack this problem.
The last point is that humans rightfully have a fear of just handing over the reins to a mechanical being. Good examples of this being:
Skynet - Computer system from the Terminator movies (left).
HAL 9000- Computer system from 2001 A Space Odyssey (right).
Believe me, you want neither piloting your ships.
I think the UAV option is quite valid, but as the challenger states, the FAA would take a long time and people would be vary wary of such a major change. The question then is: could "Sully" have pulled off what he did, without being in the cockpit of a plane?
» The Large Hadron Collider will destroy Earth. Bet: $1,000
I'll let this video do the talking for me:
| The Daily Show With Jon Stewart | Mon - Thurs 11p / 10c | |||
| Large Hadron Collider | ||||
| ||||
I shall leave you with a suggestion, if there is one thing you should do before the summer ends, go see the film "Moon".
