By Kathy Garges

[ home | newsletter | past | join | listserve | shareware | directory | links | md9 ]

Cold Hard Cash

Long ago, the center of American government moved from Philadelphia, with one important exception. The United States Mint in Philadelphia is the largest mint in the world. It produces more than 4.5 billion of the circulating coins minted in the U.S. every year and is home to the sculptor-engraver team that creates all coin designs.

Not surprisingly, robotics and artificial intelligence have replaced humans in many parts of the manufacturing process including recent advances in technology for die manufacturing and coin counting. An AI haptic arm gives coin designers feedback that feels like sculpting in clay.

Using AI in counting machines reduces the burdens of the massive amount of coin-counting that occurs every day around the world. The first requirement is quality AI image recognition. Proditec, with a facility in MLMUG territory, Havertown, PA, is one of the leading companies producing visual inspection systems for various stages of coin manufacturing. Neural networks (see the January 2013 column) are the primary AI method for counting systems, but they also use a variety of other AI methods including edge detection systems (some U.S. coin edges are reeded and countries have coins of various shapes). Requirements for imaging and counting systems range from inspection of coins for near-perfection at the end of minting to coins subject to the wear and tear of constant use, including ancient coins found at archaeological sites.

Artificial intelligence systems that detect counterfeits are also in demand. Researchers at the University of Lleida developed an inexpensive AI system to detect counterfeit euros and other counterfeit coins that uses the cursor movement guide sensor in an optical mouse. Counterfeiting is an even bigger problem with paper money because it typically has larger value than coins and can be imitated with less expensive materials. One suggestion is to embed photons in bills and polarize them using quantum mechanics. A more practical idea in the context of current technology is to create "smart money" by embedding a radio frequency identification (RFID) chip into each bill, and developers are working to overcome problems like physical flexibility and wear and tear.

Why not go completely virtual with money, using electronics as the "medium of exchange"? Currently, almost all electronic "money" transactions, like PayPal, are verified through banking back channels, re-introducing into these exchanges the problems with speed, trust and anonymity that coin and paper currency were designed to solve. Instead of being a solution, electronic money is replaying all the historical issues of currency exchange and monetary systems, and adding some new technological issues, including how to create the currency, whether to limit the money supply and whether to "back it" with something that has intrinsic value (like gold), how to aggregate micro-payments to reduce the cost of many small transactions, what to do with leftover "spare change," how to protect against potential disaster that wipes out the virtual currency, how to prevent counterfeiting, and more.

Bitcoin is an established virtual currency based on sophisticated encryption techniques with over $1 billion in circulation, but it has recently experienced dramatic price swings (critics say because it is based on the concept of a limited money supply, similar to the old gold standard). There are few merchants that accept Bitcoin. Competitors believe they can improve on the model with faster verification and a better way to create virtual coinage, but at this time virtual currency remains in the experimental stage. Not the least of the problems with Bitcoin and its competitors is the learning curve for humans to understand this new "medium of exchange."

Speaking of problems with monetary systems, some economists and politicians have seriously suggested that the U.S. mint one or more platinum trillion dollar coins to deal with the budget deficit. There is no limit on government coining of platinum money in the U.S.; trillion dollar coins could technically be minted without Congress having to raise the borrowing limit and without approval by the Federal Reserve. Other economists argue that the U.S. has effectively accomplished the same result, and could easily continue to do so, by issuing small denomination bonds to John Q. Public, who pays the U.S. Treasury for them, and then sells them to the Federal Reserve for cash.

If you do acquire a trillion dollar coin, you will automatically be a possessor of that rare property, money that's more valuable than money — or, collectible money. Collectible money includes ancient coins, rare coins, and "error coins" created in the minting process (not to worry, even with artificial intelligence the minting process is unlikely to ever be perfect). To find the trade value of your super-money, you can consult the CoinSociety's online service which continually updates the last verifiable online trade in various coins. The website's Coin Price Guide uses AI that mines online information for coin trades and classifies it into appropriate categories. To manage your hoard effectively, you will also need to acquire an AI-based coin counting machine that will run continuously in your counting house, counting out your money, while you remain in the parlor, eating bread and honey (and enjoying the MLMUG newsletter).

Sources and additional information:

Daniel Terdiman, "From metal to money: Making America's coins," CNET, July 8, 2010, tinyurl/p47xd2d

Discovery Channel, "Secret Life of Money," premiered on March 30, 2013 (look for future replays).

Tewodros Beyene Mesfin, "Modelling Banknote and Coin Recognition Systems Using Neural Network: A Case Study for Ethiopia," February 2011, tinyurl/ptme87n

C.M. Velu, P. Vivekanadan and K.R. Kashwan, "Indian Coin Recognition and Sum Counting System of Image Data Mining Using Neural Networks," International Journal of Advanced Science and Technology, Vol. 31, June 2011, tinyurl/nw38emq

"Counterfeit Euros Are Detected With an Optical Mouse," Science Daily, November 17, 2009, tinyurl/yh74663

Scott Aaronson, "Quantum Money," Fall 2009, tinyurl/o2rr349

Jeremy Hsu, "Smart Paper Makes Traceable Money Possible," IEEE Spectrum, May 2, 2013, tinyurl/ojrdysg

Salon.com, "Explaining Bitcoin," 33rd Square, April 13, 2013, tinyurl/oe8ydem

Tom Simonite, "Bitcoin Isn't the Only Cryptocurrency in Town," MIT Technology Review, April 15, 2013, tinyurl/cw4omd4

Laurence Kotlikoff, "The Treasury Has Already Minted Two Trillion Dollar Coins," Forbes,January 19, 2013, tinyurl/byq8o2b

"Coin Society.com Announces New Coin Price Guide," Numismatic Guaranty Corporation, September 26, 2009, tinyurl/p7f2mpw



Kathy Garges is a member of MLMUG.

[ home | newsletter | past | join | listserve | shareware | directory | links | md9 ]

©2013 by Kathy Garges & MLMUG
Posted 06/03/13
Updated xx/xx/13