The Arctic will never be ice free, after all, in winter, it gets no sunlight for months at a time, so even if the equatorial seas are boiling, the poles will still freeze over in winter. If the current rate of decline continues, the N pole will be relatively ice-free in summer by about the middle of this century, but I believe the rate of decline will probably slow down and the summer ice cover will level off at about 10%-20% of what it is now. Deep winter ice extents will stabilize at about half what they were 40 years ago. Complex systems under stress tend to readjust or collapse unexpectedly; but the earth has a long history of being able to adapt to change without runaway excursions into hellish conditions, I suspect the warming trend will eventually stabilize and the Earth will achieve equilibrium at some higher temperature. At least, that seems to be what has happened in the past, we won't go the way Mars or Venus did.

The summer ice cover has dropped by almost half just since we've been keeping satellite records--35 years. That's an awful lot in too short a time to just be due to "natural causes"--especially since we can't point to any natural cause that has changed that much--except greenhouse gases.

There are also a lot of negative feedback mechanisms that will kick in--for example, the warmer it gets the more water evaporates, the more water evaporates, the more clouds you get, the more clouds you get the more sunlight gets reflected back into space, the more light gets reflected back the cooler it gets. And no doubt a host of other processes are active, many of which we may not even know about. These processes tend to keep the global climate relatively stable, but if we systematically keep on adding heat, sooner or later its going to have to get warmer. There's no getting around that.

However, there is an important POSITIVE feedback mechanism at work that may partially override the negative ones. The more ice melts, the less reflection you get in summer, so the warmer the polar sea gets, which means the longer it takes to freeze over again the following winter, and so on. The effect feeds on itself, getting worse every year, especially when driven by gradually rising air and water temperatures.

I doubt the ice will ever disappear completely, all year round, but it will decrease to the point that the earth will eventually readjust to a higher temperature level. The planet will still be habitable, but the weather patterns, particularly rainfall and sea levels, will change all over faster than we can adapt to it as a civilization. Agriculture and habitability, coastal storms and flooding, all will be affected. It may even get colder in winter in some places than it is now. The blizzard winters we're getting in the NE due to a shift in the jet stream may be a part of this. So may the droughts and wildfires in the W. The snowfall in the Sierra Nevada is down to less than 20% of normal.
If this is a trend and not just a fluke, California may start to look a lot like NW Africa. We just don't know. My opinion is that the world will be different in the near future due to climate change, not deadly and catastrophic, but different. And a lot of our economic and political systems which have adapted to a relatively stable climate for the last few hundred years will be stressed by these changes. The result will be economic distress and political conflict: famine and war.

The political debate on climate change is handicapped by a lack of true understanding on what the future holds, and what to do about it But it comes down to this: should we continue screwing around with a system we don't fully understand, that we depend on critically and which appears to be changing unnaturally, simply because it is inconvenient?

Its like smoking. A lot of people smoked and lived to a ripe old age. A lot never smoked at all and still died young. But if YOU are deciding on whether or not to quit cigarettes, you should not depend for advice on the guy who is getting rich by selling them.

Here's what the melt looks like in the month of minimum ice cover, September. The graphs may look similar at first glance, but look at how the y-axis has been expanded so the entire drop can fit in one graph--from 8 to 5 million sq km in 35 years. And not only is the trend continuing to drop, the swings are becoming more violent and erratic. Put on your glasses.