人工智能vs.人工智慧!?

Yahoo奇摩理財

作者:Ralph Cheng 鄭宇和

上週末,革勢資本(Tmesis)全體成員一同欣賞了《巴黎交易員L’Outsider》。這部電影改編自真人真事:一位來自法國興業銀行的年輕交易員,因為情緒所造成的種種因素,進而犯下了金額高達500億歐元的詐欺案件。

電影栩栩如生地描繪出交易室那種高張力、高壓力的工作環境,細微地刻劃出交易員在面對誘惑、面對恐懼的各種心理狀態變化,以及這些變化如何影響到他們所做出的投資決策。在電影播放的過程中,過去同樣身為交易員的我,那些痛苦、不堪回首的回憶,如幻燈片般一頁頁地閃現在我的腦海中,與電影畫面相互輝映。

情緒容易影響交易判斷

在播放廳的燈光逐漸亮起時,我與Tmesis的夥伴們總結了對這部電影的心得:隨著交易帳戶上的金額越來越龐大,交易員的心理狀態將一步又一步地陷入不穩定的情緒漩渦,決策的品質也將大幅惡化。這樣的結果無法避免,因為這就是「人性」。上述的結論進一步解釋了為何我們要建立一個人工智能(artificial intelligence, A.I.)系統來為我們進行交易。

其實台灣將A.I.翻譯為人工智慧是非常不貼切的,因為智慧(wisdom)與智能(intelligence)有其根本上的不同,目前的A.I.在本質上仍不及智慧的雛形。

我們認為「智能」是「做出高品質決策的能力」,而最容易使交易員失去判斷能力的原因,就是情緒。有許多因素會影響並決定一個人的智能,包括過去的經驗、知識的累積、本能的直覺、及心理狀態等。這也是我們到學校受教育的目的,因為希望透過學習獲得更多知識及經驗,提升我們的決策能力。

然而,如果我們把一個交易員所做出的交易決策想像成一條數學方程式,「情緒」這個參數將毫無疑問被畫上一個大大的負號。一個擁有再高智能的交易員,只要受到貪婪、恐懼、狂喜、憤怒、征服的情緒影響,將不再適合擔任重大金融操作的決策者。

機器可以取代人類交易員?

人工智能善於消除這類型的風險,因為人工智能的系統並不會受到情緒所牽擾。

許多人聲稱,機器沒有足夠的智能,無法取代人類交易員。從《巴黎交易員》可以看到,一個從世界頂尖商學院畢業,並成為公司搖錢樹的人類交易員,因為受到情緒因子的影響而開始進行金融市場上的博弈,甚至讓公司瀕臨破產的危機。相較之下,從風險控管的角度出發,即便我們假設A.I.交易員(系統)擁有較低的智能、較少的知識、較差的直覺,也產生相對較低的報酬,但A.I.仍然比人類更為優秀,因為前者並不會因為情緒上的影響,進而做出無法挽回的決策。

在人工智能的科技不斷突破下,許多人擔心人類將會開始一點一滴地失去與機器競爭的優勢。但是,智慧(wisdom)是人類仍擁有,也最特別的一項重要資產(至少目前為止我沒聽過artificial wisdom)。

「智慧」比起「智能」要來的更難去定義,因為智慧是由各種複雜的概念所交織而成。我剖析了其與A.I.主題所有的相關特徵,並為其定義:「能夠在未知的領域中,開創並發展出新的知識和想法,這樣的能力,我們稱為智慧」。

的確在現實中,科學家和各領域專家都已展現出這樣的特質,他們致力於發展新的科技來改善A.I.在實際應用上的績效。至少就目前而言,A.I.系統並無法透過自我提升,成為更高層次的造物。

未來十年 AI革命將襲捲全球

警鐘已經浩浩蕩蕩地響破天際,人類所要面對的現實,是一波全新的浪潮,有史以來最劇烈的產業革命將襲捲而來。李開復博士在國立台灣大學畢業典禮的致詞中,提到在未來的十年,不分產業、不分年齡,所有人都必須學習如何與A.I.相處。面對新時代的來臨,每個人都必須更加靈活並具備開創新路的智慧。

文章理應隨著上個段落的結尾劃上句號,畢竟關於A.I.的論點已經闡述至一個段落、全文結構也已完整,但我的心中卻又萌出了新芽,接下來想再進一步與大家分享一些心裡的想法。如內文所述,智慧這個字有著更多也更深的涵義,而在我的認知裡,如果智慧與生命的意義有所關聯,那麼這中間的橋樑必定與「愛」脫不了關係。

舉例而言,知道如何以寬恕的口吻來回覆無知、不公正、魯莽的謾罵,是智慧;知道如何讓群體中最弱小的個體感受到溫暖,是智慧;知道如何把充滿荊棘、毒刺的世界轉變為一個更美好的地方,是智慧。這些充滿智慧的作為,都需要愛的能力,而學習如何去愛就是未來人們都必須必備的能力。我想李開復博士或許也會認同我這樣的論點,如同博士在一次訪談中分享了他的看法:未來將會有更多的職業是關於傳遞經驗和愛。

在文章的起頭,我並沒有預期到最後這樣哲學性的發展,但希望你們都喜歡。

更多文章請見法意專欄好文

Intelligence versus Wisdom

Last weekend, our whole company went to see Team Spirit (French: L’Outsider) together. The movie is based on the true story of a young Paris trader who was involved in one of the biggest financial fraud cases in history. It vividly depicts the different psychological stages and the high-stress working environment traders experience. For this reason, it evoked many of our own painful memories from our time as traders. At the end of the movie, we concluded that as the size of a trading account grows, the trader becomes less emotionally stable—and the quality of the trades he or she makes deteriorates. This is an inevitable result of human nature.

This conclusion reinforces the reasons we build artificial intelligence (AI) systems to trade for us. Intelligence is defined here as the quality of decision-making, and traders usually lose a lot of their intelligence when the effects of emotions kick in. Several factors determine and affect one’s intelligence. These include one’s experiences, knowledge, natural instincts, and psychological state. Accordingly, we attend school to acquire knowledge, gain experience, and, hopefully, leave having enhanced our decision-making abilities.

However, if we conceive of the quality of a trader’s trading decisions as a math function, emotion would definitely be marked with a huge negative sign. Greed, fear, ecstasy, and anger can easily compromise what might originally have been the high intelligence of a person and can quickly make the person unsuitable to make important financial decisions.

Artificial intelligence can perfectly minimize these risks because AI systems are not affected by emotions. Many argue that the machine is still not smart enough to replace humans in the trading room. However, from a risk management perspective, I would argue that even if an AI trader has lower intelligence, less knowledge, and worse instincts than a human trader, and these result in it yielding only trivial returns, it is nevertheless the case that the AI trader is better than a human trader who went to the most prestigious business school in the world but whose intelligence diminished because of the aforementioned factors, leading him or her to gamble and bankrupt his or her whole company.

Amid the ever-increasing presence and advancement of AI technologies, many are worried that humans will lose all the advantages they have over these machines. However, there still remains a remarkable quality that only humans possess—wisdom (I, at least, have not yet heard of artificial wisdom). Wisdom is much more difficult to define than intelligence.

Because it is such a complex concept, I will try my best to first define only the characteristic of wisdom most relevant to the topic of AI: the ability to develop new knowledge and thoughts about uncharted areas. Indeed, scientists and professionals exhibit this very quality as they eagerly develop new technologies to enhance AI performance. AI systems, at least for now, cannot build themselves into higher beings.

Nevertheless, the reality remains that the alarm signaling a new wave of industrial revolution is sounding louder than ever. As Dr. Kai-Fu Lee mentioned in his speech at National Taiwan University’s graduation, people, no matter which industries they are in, must learn to get along with AI in the next ten years. To prepare for these developments, one must become not only smarter but also wiser.

As I mentioned in the previous paragraph, there is much more to the word wisdom. Here, I will provide a more personal definition of it. In my opinion, if wisdom concerns the meaning of life, it must be closely related to love. For example, knowing how to reply to ignorant, unfair, and rude bad-mouthing in life in a forgiving tone is wisdom; knowing how to let the weakest among us feel the warmth is wisdom; and knowing how to transform the vicious world into a more beautiful place is wisdom. These actions all require the ability to love. I think Dr. Kai-Fu Lee will probably agree with me that love is a quality humans will definitely need in the future. Indeed, he shared in one of his interviews that future careers will be about spreading love and experiences.

I did not expect my article to become so philosophical at the end, but I hope you liked it.

★延伸閱讀★
買基金?買ETF?巴菲特真正的想法?
交易層面的先進量化與落後量化區別原則
計量金融分析師如何看待升息這件事?
法國大選中的國家陣線

【作者簡介】Ralph Cheng 鄭宇和

海外量化對沖基金CIO首席投資官
畢業於西雅圖華盛頓大學 University of Washington – Seattle
Acquisition International 2017 東亞最具創新基金經理人
Acquisition International 2017 年度最佳對沖基金

 

分享連結

財經新聞

  • 重大要聞
  • 國際財經
  • 大陸香港
  • 基金動態
  • 個人理財