"This paper presents a new formula for calculating when fossil fuel reserves are likely to be depleted and develops an econometrics model to demonstrate the relationship between fossil fuel reserves and some main variables. The new formula is modified from the Klass model and thus assumes a continuous compound rate and computes fossil fuel reserve depletion times for oil, coal and gas of approximately 35, 107 and 37 years, respectively. This means that coal reserves are available up to 2112, and will be the only fossil fuel remaining after 2042." http://www.peakoil.net/publications/when-will-fossil-fuel-reserves-be-diminished

    Current trends cannot be continued until 2100. There isn't enough fossil fuel. All the easily reachable oil is going to be burned and go into the atmosphere, no matter how successful the attempts to curb global warming. If we dramatically curb the use of fossil fuels, it will simply take a little longer to burn it all up. But it is all going into the atmosphere. There is no political force strong enough to tell all the people of the world they can't run their cars or heat their homes with fossil fuels any more. Since is entirely futile to stop all the easily extracted fossil fuel from being burned, there is no point in debating how to curb emissions. Our focus should be on what we can do to minimize the impact.

It is kind of funny how the marketing departments of big software companies think we actually look forward to 'updates' which annoy us and waste our time. Now I actually breath a sigh of relief when my BlueRay player gets past the moment where it may insist I have to spend 5 minutes 'updating' before I can watch my movie. I can't imagine wanting to pay $100 in return for being hassled with updates I don't care about. Apparently, they haven't figured out that people very well might pay $100 to never be bothered with them.

You are absolutely right I did not use the correct word, but I still am not interested in searching for pages which have copied the text from Wikipedia without adding anything original.

My pet peeve with google searches is when I get page after page of pages which have just stolen the text from Wikipedia and placed it on their site with ads.

Who is responsible for the AI's actions? Is it the machine? Is it the person who setup and turned on the machine, or the person who designed the machine?

        As far as the law goes, isn't it important that the accused understand their own actions? I.E. the insanity defense allows you to prevent taking responsibility for your actions. So if the machine does not understand anything, then how can it be held responsible?

      By this test, the responsibility for the non-self aware machine's actions should lie with the person who sets up and actives it. They should understand what the machine may do when they set it loose, autonomous or not.

      Then the comes question, what does it really mean to 'understand' something? I don't think anybody yet has a good answer for that, but no machine I've heard of seems to actually understand anything yet. I think this has been a big problem in the pursuit of the self-aware machine. Nobody understands understanding yet, so how can they build a machine which does it?

      In the article, they mention how robots are immune to fatigue or emotion etc, but I think that may not necessarily be the case with a self-aware AI. Are we so sure that you can make a self-aware machine without emotion or feeling? Isn't it ultimately a balance of conflicting emotion that drives us? How can you *want* to do something if you do not have any feeling? And if you don't *care* about anything, then there is no reason to any action over any other. An intelligence that does not care about anything may just as likely commit suicide or sit like a lump rather than something useful to itself or us.

    I think the truly intelligent machine is coming, but I don't think it operate at all like a drone with logic. When you get a machine which is self-aware you not only have ethical issues about what the machine does to humans, but also about what humans do to the machine. That may sound silly, but I can imagine that a self-aware intelligent machine with feelings and memory could be very vulnerable with no ability to defend itself from its owner.

Density and cost may be the holy grail, but there are a number of other properties which important which can make the difference between marketplace success or failure.

1.) Data Reliability
2.) Data Lifetime
3.) Store element lifetime - do the individual storage locations 'wear out' like flash?
3.) Read/Write Speed - Although there is an article in Science Daily which reports 100ns
4.) Power Requirement

I didn't see any of these mentioned in the article. If these properties are all equal to or superior to flash, it could be the next big thing.

At least it didn't get on the radio and swear at everybody, then grab two beers and jump out into space.

Or maybe it did....

You can still use integer math to represent fractional values. For example, using the upper 16 bits as the integer part, and the lower sixteen as the fractional part.

    Something like this only implemented with inline assembly:

Int32 fMult( Int32 a, Int32 b)
{
      return (Int32) (((Int64) a * (Int64) b)>>32);
}

You don't have nearly the dynamic range of floating point, but you *can* implement rotation matrices, vectors, time and distance and physics calculations. You just have to be careful to keep the values in range.