We've all heard the stories... millions of embedded controllers will fail, computers and communications down, no power, civilization will fail and so on. No offense, but all that is over sensational crap.
Worse case scenario
Let's be serious for just a minute. Worse case scenario, a large fraction of power systems (and utilities) fail, it gets hard to figure out who's money is whose (short term), some gas stations don't work, a few banks have problems, and so on -- what is the big deal? Have you seen cities hit by earthquakes, floods, hurricanes? Those do far more damage to the infrastructure -- and humanity and civilization not only survives, but people tend to band together to get through it. Usually those disasters bring out the best in people -- not as much of the worst. So why would this be any different?
Even Worse - Even if the problem is 100 times worse than we think, people aren't out on the streets. Landlords aren't going to throw you on the street the first day you can't pay (because they have the same things to worry about as you do) -- worst case the economy will be put on hold while things get worked out. Food will still be distributed, you can boil water. We may have long lines at banks while some have to go back to paper-trails and written records and so on, but that is about it. Likely companies will need to hire more help (temporarily) to get things sorted out and handle things manually. Power companies can redirect power -- and all the delays will be temporary and usually pretty spotty. I personally think civilization can get through that with barely a hiccup. There will be some scams and attempted fraud, and some profiteering, big delays and kinks in the system. Society can survive not much getting done and a far bigger step back in efficiency that is likely from Y2K -- society will go on.
Back to reality
Now the reality is that things are unlikely to be anywhere near that bad. We've been aware of the problem for decades. Most companies have overreacted and over engineered the systems. Many have tested and have trial runs. Consultants have been hard at work (well, for consultants), and getting paid for fixing the bugs (that many of them helped to create). How much is going to fail? We don't know -- but odds are it won't be as bad as the nay sayers predict, but there will be some unexpected surprises.
Microcontrollers are not the big deal that some pretend. Chicken Little's talk about millions (or billions) of microcontrollers in various embedded systems -- that is true. Many can't handle 4 digit dates -- that is true too. Most don't handle dates at all because it is irrelevant to their jobs. Does your car or toaster need to worry about the date? I bet the gross majority of microcontrollers just won't care at all. The gross majority of Microcontrollers are in very simple systems like microwave ovens, VCRs, refrigerators, and so on, and if they fail, other than a moan, it is no big deal. Quite a few of the remainder have been hunted for (if they were in critical systems), and tested to see if they would fail. So what is left?
The real question is not how many Microcontrollers are out there. It is how many critical systems that did have a requirement for knowing the date, that the companies don't know about, that companies haven't tested for (and failed to find in trial runs), that a programmer was stupid enough to use a 2 digit date for, and that will have an issue when it rolls over to zero, that people won't be able to find an fix or identify immediately? Probably not very many because that is a pretty fine net to get through.
Contrary to hype, not all systems are interdependent -- in fact most are not. Many critical systems have contingency plans and ways to route around a problem if one should arise. Companies may lose track of a little billing, or have troubles tracking you down later, but they can just give you service for free for a while and survive. There are layers of options. So Phones, Power lines, medical instruments are tested, and stand a high likelihood of working -- the things most likely to fail are the things least likely to impact society in any serious way.
There are a lot of corporations that existed for hundreds of years without all the computers and automation. Going back would be painful, and cause short term losses, and lots of new temporary hires while fixing things, but is unlikely to wipe many out -- or at least it won't wipe many out that aren't already having many problems. The new workload (demand) required by some companies having problems is more than likely to outpace added dumping of supply (from the few corps. that crumble) -- and often a loss for one company is an opportunity for another, and other companies will be snapping up sold off people and divisions and so on. There has to be a HUGE and very sudden impact before things go critical -- and I don't think we have anything close to that. Companies are doing dry runs as well -- but aren't as bright (over-all) as mission critical institutions (like utilities) but again, it just doesn't seem that big a deal.
Banks have been doing dry runs for years. They have put the most money into making sure they don't lose track of their money. There may be some difficulty transferring money bank to bank as easily for a short while (especially out of country) -- but I think that will be about it. Banks already have paper trails and redundant systems to avoid the loss of money -- so they are probably the safest (overall) -- and failures and losses are covered by Govt. promises to print more money (i.e. insure your money against loss). There will be some opportunities for fraud, and misappropriation, but most of the time people won't know how to exploit them until it is too late (or they are fixed). So I just don't see the collapse of anything as a big threat.
The bigger threat is to other countries where they haven't taken this stuff seriously AND they've been buying all our legacy systems (that are more likely to have problems). Remember, the U.S. sort of subsidizes the world by selling off our surplus to other countries (and getting rid of our problems). Many companies have already replaced systems that may have problems -- and some have snapped them up (for pennies on the dollar) and may get bit. But those countries are also a lot less dependent on technology -- so the pain of failures is far lower.
Remember the people making the biggest stink about the Y2K problem are the ones with the most to gain -- the brokers playing the market, the press getting lots of readership, the consultant who get paid to talk about or fix the problems.
So all this bullshit about any single system / failure bringing the world to its knees is excrement. The survivalists will come crawling down out of the hills looking stupid. It will probably cost some companies a little bit in billing and added overhead, which they will get around just like a small recession. But in fact, there is more potential for a boom market -- as quite a people replace their few VCR's and Microwaves that stopped working, and more people are hired to fix and audit paper trails. Not to mention the raised consumer confidence (and feeling of victory) since nothing serious happened. Less unemployment + more goods being bought + consumer confidence restored = good economy. If other countries are having more problems then the U.S., that means the Dollar value is raised (other monetary unity devaluated) and the purchase power of American increases -- and a better economy (for Americans).
The only real fear is of fear itself. If enough people THINK the world is going to end -- then we get a bunch of people doing stupid things (like pulling all their money out of banks, the press sensationalizing it, scaring more people, and causing some runs on money, etc.) -- but this isn't the 1920's. We have more systems in place, and less reasons to panic. Market-makers will play games (with hype) and want to scare people and rock the boat -- the waves give them opportunities to make money. And there may be a small amount of economic turmoil -- but I'm saving now so I can invest if others start bailing out (remember buy-low, sell high). I think there are going to be a few opportunities, a few losses, and in a few years (say 5 - 20) no one will remember what all the hype was about. After tension is removed, the economy usually thrives -- and frankly considering purchase rates and things like that, I don't think there is that much tension out there.