The Turing Test is meant to gauge a machine’s intelligence. The test, as proposed by Alan Turing in 1950, asks for computers to imitate human beings well enough as to believably carry on a conversation with a human, such that the human does not realize he or she is conversing with a machine instead of an actual person.
The Loebner Prize is an annual competition that presents a platform for teams and their chatbots to see how they fare in such an imitation game and to ideally pass the Turing Test. The winner of the Loebner Prize is that bot that is voted to be the most human-like computer.
Brian Christian participated in the 2009 installment of the competition. He did not contribute a chatbot, but rather was one of the human confederates. Just like a software bot, the confederate’s task is of course also to convince the judge of his humanity during their written chats, thus trying to keep them from judging a computer program to seem more human-like than him based on a conversation. As it turns out, being the most convincing human human, has its rewards, too:
But there is also, intriguingly, another title, one given to the confederate who elicited the greatest number of votes and greatest confidence from the judges: the “Most Human Human” award.
One of the first winners, in 1994, was Wired columnist Charles Platt. How’d he do it? By “being moody, irritable, and obnoxious,” he says – which strikes me as not only hilarious and bleak but also, in some deeper sense, a call to arms: How, in fact, do we be the most human humans we can be – not only under the constraints of the test, but in life?
An intriguing question indeed! Competing against software that strives to be as human-like as possible can serve as great motivation to contemplate what exactly it means for a person to come across as a human – other than just being oneself.
Brian Christian’s book The Most Human Human: What Talking with Computers Teaches Us About What It Means to Be Alive examines that question in some depth. Our notion of (artificial) intelligence and valid tests thereof keep changing as computer become able to accomplish tasks that were previously assumed to take real, human intelligence. Chess was a great example of this and so was the game of Jeopardy.
As computers and our capacity to program them and make them smarter improves, the machines appear to be gaining ground. Does that mean it is just a matter of time, until the machines will pass the tests we present or are we able to improve ourselves to stay ahead of them? The author seems to think so:
In an article about the Turing test, Loebner Prize co-founder Robert Epstein wrote, “One thing is certain: whereas the confederates in the competition will never get any smarter, the computer will.” I agree with the latter, and couldn’t disagree more strongly with the former.
The author joined Jon Stewart for a brief segment on The Daily Show to discuss his book, the Loebner Prize and Artificial Intelligence:
It is a brief, but informative conversation. My favorite part occurs around the 2:35 mark. Jon Stewart: “Tell me, how computers have progressed – they’ve been able to, obviously, beat us at chess, and now at Jeopardy … Will they move on … beyond our hobbies? [... or will they always be stuck in these types of games in their capacities?]”
The Most Human Human is a thought-provoking, engaging read – highly recommended.
Does This Headline Know You’re Reading It? discusses Text 2.0, a very interesting AI research project that focuses around this premise: What if your computer knew, what you are reading, as you are reading it?
This is an intriguing question and this work could lead to a multitude of interesting applications. The following video (it starts a bit slow, but includes interesting examples later on) shows just a few of those.
People are getting more and more used to rely on technology and getting walking or driving directions using their GPS devices in the car or their mobile phones. I know few people who prefer asking for directions to using their phones in an urban setting.
Autonomous City Explorer (ACE) on the other hand is a robot that has to rely on successful collaboration with humans. According to thefollowing video ACE successfully navigated from the origin to its destination (a distance of about 1.5 kilometers) in about 5 hours – and by asking 38 pedestrians and interpreting their hand signals along the way.
This projects seems like a nice venue to explore a big collection of different problem areas, such as collision avoidance, gesture interpretation, detecting of people, route planning, and many more. I wonder what this project will teach us about the nature of collaboration itself, particularly between humans and machines.
Imagine this situation: A company has hundreds (maybe thousands) of employees. All of them have their own skills and areas of expertise. There is probably lots of overlap, however any one person will not know everyone in the larger group who has particular skill sets. It someone is working on a project and needs assistance to overcome some technical hurdle, it could be very helpful, if they could communicate with those people who also have experience in that area. Those people might be located in entirely different parts of the company.
Email addresses are a means to an end. The goal is usually not to send an email to a particular address, but to a particular person. You want to say hello to your friend Steve or send a message to the VP of marketing at Microsoft or to the head caterer for your wedding. Ideally, you could send a message to a person just by entering his or her name, position, or some other descriptive attribute. If a person’s email address changes, the email system should send to the new address automatically. If the person matching a description differs over time, the email system should send to the person currently matching that description.
In the given example, the user would be able to get answers to his or her questions by reaching out to the people with the fitting skill sets without previously having known those people: The email system can decide, who the most appropriate receivers of the messages are.
I cannot help thinking that Aardvark was at least a little inspired by the ideas behind semantic email addressing. Their process is simple: Users send in questions (using email, twitter, IM, etc.), Aarvark routes the question to another user is (hopefully) qualified to answer it and the user will eventually receive a response, often just a few minutes later. In this social search approach, Aardvark accomplishes the job of finding information by finding the right people who can provide it. The service has received very good press and was recently acquired by google.
Twitter seems like it might be a good platform for this problem area. If someone has a public twitter feed, they are essentially broadcasting their updates to the open stream and anyone can see them. It is probably safe to assume, they are at least open to the idea of talking to strangers/responding to messages from people they do not already know.
How could one go about finding the best people to message though? One method is certainly to search the message stream for specific keywords and basically manually look for people who might be active in areas of interest. You can also search in and add yourself to oneofthemanydirectories that are being developed.
But, if I simply need to talk to someone and ask them “May I ask you a question about XYZ?” then clearly, a) broadcasting my question hoping that someone will answer could be very inefficient and b) first researching who the best person might be for my question(s) puts all the burden on me.
What if the user could simply send out the question and the system would ensure that the most appropriate people see it?
The basic idea here is this: The user submits the question (along with a set of keywords) to his or her software. The software has analyzed other users’ message streams, extracted keywords, etc. and generated a knowledge base. If the query can be confidently matched to another user, a message is generated and send to that user. The message will be visible to that user as a regular name mention and they can choose whether to engage in that conversation.
Some of the obvious challenges:
Generating of meaningful keywords/subject areas based on a person’s message stream.
Successful matching of queries with users.
Establishing an effective communication protocol that does not easily lend itself to abuse, i.e. spam.
A lot of web-based social networks are great at helping you connect with people you already know. Twitter makes it easy to connect with new people. The outlined approach (or a variation thereof) might be a good way of further supporting creation of those new connections, based on areas of interest.
