随着中国最年轻的教授级研究员在中南大学诞生，刘路教授级研究员名震华夏，而和刘路教授一起成名的是一个大家闻所未闻的猜想——西塔潘猜想。

      因为陈景润、佩雷尔曼，国人最熟悉的猜想当属哥德巴赫猜想和庞加莱猜想。而提出这两个著名猜想的哥德巴赫和庞加莱都是数学大师。尤其是庞加莱，被公认是十九世纪后四分之一和二十世纪初的领袖数学家。

      现在由于刘路教授，“西塔潘猜想”也进入到大家的视线了。“西塔潘猜想”到底是个神马东东呢？属于著名猜想吗？下面我首先来介绍一下西塔潘猜想，顺便八卦一下刘教授的工作。

    刘路（刘嘉忆）的论文名叫“RT_2^2 does not imply WKL”，为了理解他的工作，首先要知道“RT_2^2”和 “WKL”分别代表什么。
“RT_2^2”代表一个定理：假设有一个无限图，它的顶点对应全体自然数，任何两个顶点有一条边，现在对这些边任意二染色，该定理断言，无论怎么染色，必然能找到无穷个顶点，它们之间的边都染同一种颜色。这是无限拉姆塞定理的特例。
    “WKL”代表另一个定理(Weak Konig Lemma)：每一棵无限的二叉树必定包含一条无限长的分支。解释一下：二叉树是指每个节点最多有两个儿子节点的树；无限二叉树是指该树有无限个节点；包含一条无限长的分支是指存在一条路径，从该树的根节点往子节点走，能无限地走下去。这个定理很好理解，不难想象如果二叉树的任意分支都有限长，那么树的节点必定是有限的。
    刘的工作顾名思义，就是证明定理“RT_2^2”推不出定理“WKL”。所谓的西塔潘猜想估计就是指猜测“RT_2^2能推出WKL”。刘路的工作的PPT见liulu.pdf

我再来介绍提出这个猜想的西塔潘本人吧，西塔潘是否也是一位学界大师或名人呢？偶尔看到一篇1999年的文章介绍西塔潘先生。

西塔潘（David Seetapun）1995年在加州伯克利获得数学博士（数理逻辑领域），他的导师Ted Slaman对西塔潘有很高的评价。西塔潘毕业后就离开了伯克利，先到瑞士银行，后又去了高盛伦敦分行做期权交易。由于他的数学才能，两年后西塔潘的年薪就达到了百万美元。但很快，西塔潘建立的数学交易模型整体崩溃，银行损失近1亿美金，西塔潘理所当然被投行解雇。那么解雇后的西塔潘去哪里了呢？

     解雇后的西塔潘离开伦敦去了著名的赌城拉斯维加斯。在赌城，he played the blackjack tables。后来，他又移居到佛罗里达，在那里从事一份很危险的深海船员工作（where he has the dangerous job of a crewman on a deep-sea swordfish boat）。

也许令西塔潘万万没想到的是，他在加州伯克利读书时提出的一个猜想，20年后成就了中国一个最年轻的教授级研究员 。

The postdocs who lost millions

http://www.timeshighereducation.co.uk/story.asp?storycode=148896

19 November 1999

THES Editorial

David Seetapun and Andrew Felce gave up glittering university careers to trade for huge salaries at leading investment banks. Both lost their jobs as a result of trading losses.

British-born Felce moved first. In 1992, he completed a PhD in string theory at Princeton University and became a postdoc at the University of California, Santa Barbara. Both institutions boast top physics departments and Felce was expected to quickly break new ground in his subject. Instead, in 1994, he left UCSB for an option traders' desk at Bankers Trust.

Even more was expected of David Seetapun, who gained a PhD in mathematical logic at the University of California, Berkeley, in 1995. Seetapun's adviser Ted Slaman says: "In his identity as a mathematical logician, David is brilliant. Our joint paper includes his ingenious solution to a well known and notoriously difficult problem and a perfectly routine calculation of mine. I wrote it to make his theorem available to the rest of the mathematical community."

In early 1996, Seetapun left Berkeley for the bank Credit Suisse. A few months later, he and Felce were headhunted by the London branch of investment firm Goldman Sachs. They specialised in trading options linked to interest rates. Mathematical wizardry was needed to handle these instruments, which involved not a single underlying asset such as a share price, but a continuous curve of interest rates with differing maturities.

Seetapun headed an aggressive proprietary trading venture intended to compete directly with the likes of LTCM. His reputation was widespread, and in March 1998, he was headhunted back to Credit Suisse. He was earning $1 million a year including bonuses. At Goldman Sachs, Felce took over Seetapun's old trading positions.

In September 1998, everything fell apart for Seetapun. The models stopped working, and he lost almost $100 million, before he was finally dismissed. Felce's dismissal came in August 1999, during another bout of market turmoil.

Goldman Sachs has admitted losing $20 million on Felce's trades, although traders at rival firms insist that the true figure is higher.

Felce is now unemployed and living in London. Friends say Seetapun left London for Las Vegas, where he played the blackjack tables. Next he moved to Florida, where he has the dangerous job of a crewman on a deep-sea swordfish boat.