hexagonc

[EDIT] And as an addendum to this fun tangent, even if the AI was capable, in principle, of reading billions of webpages fast and well, we don't know the power requirements would be for this. This hypothetical small company may have stumbled upon the right algorithms and the right training data to produce a true AGI, but they may simply not have the hardware or the engineering know-how to scale it up across multiple processors. Or scaling it up may require too much (i.e., more than the company can afford) data bandwidth if the robot is controlled. Again, it depends on the details of how they arrived at the AGI.

[EDIT] - My point here is that even if improvements to battery technology remain slow or constant, with sufficient innovation in energy efficiency and algorithms, we could still have robots with usable to good running times. Just figuring out a reliable walking gait and building materials for passive dynamic walking can significantly reduce the power requirements.[3]

[1] - http://hypertextbook.com/facts/2001/JacquelineLing.shtml [2] - http://www.chem.hawaii.edu/uham/bat.html [3] - https://en.wikipedia.org/wiki/Passive_dynamics

Technically, Reuters could have gone to court on her behalf but why bother unless she was some highly sought-after journalist? Intimidating letters are meant to make you weigh your choices carefully. Also, while Law360 didn't literally sue her, it gave every indication that it intended to if she continued at Reuters.

Unfortunately, this isn't quite accurate, as shown by the Q&A at the end of the article. Although her previous employer didn't directly sue her, it essentially threatened to sue her. Moreover, the way her new employer found out about her previous non-compete agreement was because her old employer specifically notified them and indicated that she had confidential information. Her new company basically had no choice but to fire her.

Even if the person who is operating the device doesn't know if the grenade is present, they may be able to pick up unconscious physical tells from the person who does (if he is present during the test). When the person who knows sees the tester approaching, they may unconsciously reveal this knowledge through body language that could be picked up by the tester. A similar phenomena occurred with a horse[2].

[1]https://en.wikipedia.org/wiki/Ideomotor_phenomenon [2]https://en.wikipedia.org/wiki/Clever_Hans

The key thing to take from this is that there is a huge gulf between the best robot competitors and the worst. The best robots seemed like, with more hardening, they could be deployed in certain disaster areas even today. The worst robots seemed to struggle just walking on two legs on anything but perfectly smooth terrain. Laugh while you still can at the numerous fail videos that will emerge but the KAIST team's robot will make you a believer. That robot was very clever as were most of the other robots from the top teams.

A slight correction to the linked article: the IHMC robot did not just promptly fall after climbing the stairs and celebrating; instead, its arm got caught on the rails of a platform that the stairs led to. It had already completed all tasks of the course.

There are no toy gas powered helicopters. The idea that they are toys are what gets people in trouble in the first place. Even small electric palmtop RC helicopters can cause significant injury if you accidentally fly it toward your face.

I think the reason this isn't done is because Wolfram Research would have to really take responsibility for calling this thing a programming language. They'd have to publish their sources or at least the requirements for what qualifies as a valid source. They'd also have to publish their curating methods so that anyone with access to their sources and the language specification could produce identical results to identical source code. Basically, their position would be no more privileged than Microsoft's with respect to CSharp and the .NET framework. I don't think Wolfram Research wants to face a Mono-like competitor for their language and that is why portraying the language as some general purpose tool for everyone rings hollow.

In other words, a computer has to actually model the world and the changes to the state of the world as the conversation goes on in order to pass a Turing test. I didn't read anything in the original description to suggest that was happening.

I'm currently working on a system for understanding natural language and you might be surprised by how many assumptions one must make in order to understand just this simple sentence. The only part of this sentence that nearly all humans could understand is the concept of growing up. Everything else would have to be inferred inductively from our personal expectations and experiences. For example, from this sentence, you wouldn't know the gender, race, or even species of Jane except from your experiences of people saying similar things about people about of whom you already knew the race, gender or species.

Consider the difficulty of modeling that sentence in a computer system. If all you had were examples of texts to go by, such as the scenario in the grandparent, how would you determine that Jane was a female? Within that vast body of text, somewhere, it has to contain the statement that "Jane" is a female name. Or Jane can only appear in sentences as the referent of a feminine pronoun, like "she" or "her". Or in that vast database of images and video, all human beings that were identified as Jane have to have female characteristics.

From an even deeper philosophical perspective, how do you know which Jane this is referring to? Is this Jane supposed to be an actual person with hidden but unknown state (such as the name of her parents) or is this a purely fictional creation, for whom it would be meaningless to ask who her parents were? How do you teach a computer the concept of a fictional character? The interesting about fictional characters that they do not describe what something is, but instead, describe what something is not. In order to create a computer that could pass the Turing test, the computer would have to be capable of modeling both fictional and concrete things at a minimum. It would have to know when someone was talking about something fictional or something that is supposed to represent an actual object. If we do not destroy ourselves first, the day will come when computers will be able to make this distinction, but I think the design of such a computer will have to be evolved rather than architected from the top down. I think the problem is just too hard.

[1] http://en.wikipedia.org/wiki/HP-48_series [2] http://en.wikipedia.org/wiki/Conway%27s_Game_of_Life

Your problem seems more with badly constructed public apis rather than overriding itself. The only public methods of a class that should be overridable are those with a simple and well-defined expected behavior. Overriding doesn't have to imply tight coupling. So long as it adheres to some expected contract (e.g., input/output ranges and exception handling), an overridable method should be just as much a black box to the parent class as the subclass. An overridable method with no side-effects is equivalent to an initialization parameter to an encapsulated object.

* Protected methods suffer from a similar problem - what kind of api is public enough to allow access by a subclass but not public enough to allow access by a wrapper?*

You have to think of the consumers of the class. Consumers that only need to use the class and are satisfied with the publicly available interfaces only need access to the public methods. Consumers that want to change the behavior of the class (often for the benefit of other consumers) will override the protected methods. Protected methods offer a set of extension points to consumers while providing useful default behavior for those that do not need them.

I'm not sure at all that it is clear what the nature of that link is. I have no insight into the gmail's email relationship modeling but I suspect that it doesn't automatically build a relationship between david@gmail.com and other people that onion2k contacted. For one thing, it would be impractical to do this across all possible relationships -- there would be polynomial growth of the adjacency matrix with no information gain to justify it. Ask yourself, what does it mean that david contacted onion2k and onion2k, at some point in time, contacted alice, bob, and charlie? Unless you already know something nontrivial about these people, you assume they are independent.

I guess my point is that data analysis can be very costly; it is reason that humans have a hard time dealing with meaningful relationships spanning more than about 150 people[1]. Even a company with Google's resources would be overwhelmed if it had to store and take into account a large number of mostly (I mean, vastly mostly) meaningless relationships. At the least, it takes away resources from the relationship graphs that do matter and which are actionable, such as the graphs of spammers.

[1] http://en.wikipedia.org/wiki/Dunbar%27s_number

[1] http://en.wikipedia.org/wiki/Nausica%C3%A4_of_the_Valley_of_...