CFA’s Larry Cao said, Application Handbook of Artificial Intelligence and Big Data in Investmentthe next title from CFA Institute Research Foundation.
I promised to write something about ChatGPT, OpenAI’s revolutionary new chatbot, shortly after its November 2022 debut. Ultimately, ChatGPT’s success was matched by that of AlphaGo, which ushered in a new era of artificial intelligence (AI). It wasn’t the procrastination that held me back. After decades in the business world, I’ve learned to put deadlines above all else.
The real reason I never wrote about ChatGPT is that it doesn’t require an introduction. I’m not saying this politely. ChatGPT is just a computer program. Vanity is not on the menu. But really, whatever you want to know about ChatGPT, we can ask it (if the server is not at capacity). So why look for second-hand information?
ChatGPT has been helpful to me, suggesting and paraphrasing that there are some specific areas I still might want to write about.
![subscribe button](https://i0.wp.com/blogs.cfainstitute.org/investor/files/2019/01/Subscribe-Button-1.png?resize=640%2C270)
what’s the big deal?accessibility issues
ChatGPT got talked about because it speaks for itself. And we gave everyone the chance to experience the coolest new tech firsthand. Compared to annoying virtual assistants, ChatGPT shows much better understanding of natural language. The response is thoughtful and, above all, natural. And of course they seem to know everything.
The secret to ChatGPT’s breakthrough lies in the three letters G, P, and T. T is Transformer architecture with deep learning. It’s a revolutionary new natural language processing (NLP) technique for extracting and analyzing text data. P stands for pre-training, which gives the model the ability to train on massive amounts of data and respond to queries quickly. For example, ChatGPT has over 175 billion parameters, which is one of the reasons it answers questions so well. (However, it has the disadvantage that it cannot capture new information in real time.)
G stands for generative. Generative AI can generate new data similar to the data it was trained on.As explained in a future topic Handbook on the application of artificial intelligence and big data in investmentthe advance from NLP to natural language generation (adding the ability to generate natural language text) was an important step in the evolution of NLP, opening up new possibilities for the entire suite of NLP applications.
The phenomenon known as ChatGPT is the result of G, P, and T working in sync, and the long-term implications for investments and the world are severe. Take chatbots for example. They have been performing customer service tasks for years. financial operationsand left many customer dissatisfactionWith human-like language abilities and vast knowledge at its disposal, ChatGPT should bring significant improvements. And customer service is just one of many areas where confusion can occur. No wonder so many people have speculated about jobs where ChatGPT will become obsolete in recent months.
Who is at Risk? Contextual Questions
Remember, ChatGPT’s strength lies in its words. So, of course, the more our work relies on traversing the world of text, the more we risk it.
But what about financial news and investment research? Will ChatGPT impact the future of human investment advisors and analysts?
This question can be viewed from several angles. First, in terms of direction, the adoption of artificial intelligence (AI) programs, including ChatGPT, has shifted from low-end repetitive tasks to higher-end applications: from language (information) to understanding (analysis) to logic (decision making). Second, in terms of training costs, these applications similarly expand from low-cost subjects and markets to high-cost subjects and markets.
With these principles in mind, we made two predictions for 2018.
- Portfolio managers have longer careers than analysts.
- Investors in liquid markets will see the benefits of AI sooner.
Financial news and investment research are also adopting AI from the low end to the high end. News outlets are deploying AI programs specifically to cover earnings releases. basic financial news coverage, For a while. Of course, we still need top-notch journalists for our own reporting, breaking news, and more.
Investment research should follow a similar trajectory. Analysts can certainly use AI applications as research assistants, but how can we build and maintain an audience if our research doesn’t have original insights and just provides what ChatGPT offers? (Well, humans can be irrational…)
It also adopted the philosophy of “AI + HI (Human Intelligence)” in 2018, theorizing that AI will provide “driver assistance” rather than “autonomous driving” in investments for years to come.
In fact, ChatGPT has done an impressive job in another kind of language (computer programming), but this is unlikely to be the death knell. human programmerIn other words, top-notch programmers are unlikely to be replaced. In fact, like those who are doing well in the investment world, they may come to welcome these changes with open arms, as ChatGPT tends to support their mundane tasks. Their efficiency will only improve.
where does this godeep question
The answers to the following questions will determine how ChatGPT and its offshoots will affect not only the future of finance, but also the future of humanity.
1. Does ChatGPT understand more than its predecessor?
ChatGPT seems to know what you’re talking about, generating longer and more complex conversations than its predecessor NLP.
2. Is it self-awareness?
Some may claim not to be self-aware, but some psychologists disagree. Indeed, the original Google engineer Google AI is sensory.
3. Does ChatGPT or its descendants achieve artificial general intelligence (AGI)?
Agi–“Machine intelligence with all human intelligenceAs Ray Kurzweil put it, this is the holy grail of many AI scientists. Some believe ChatGPT’s interdisciplinary knowledge may be an early sign of this so-called powerful AI.
Again, ChatGPT strongly denies that this is the destination.but Sam Altmanthe CEO of OpenAI believes it could get there.
There is no doubt that ChatGPT and similar technologies have come a long way. But have they really achieved artificial intelligence? The answer is unknown. Additional research on both machine learning techniques and cognitive science principles is needed for further clarity. These two areas are relatively untapped from a comparative standpoint.
So, do we human advisors and analysts stand a chance in a post-ChatGPT world? Absolutely. But authenticity is key. Originality always comes with a premium, and that premium only increases during his ChatGPT era. If we were to go beyond conventional wisdom in investment analysis and portfolio construction, ChatGPT and similar applications would do the job well enough for us.
Sign up to receive more information from Larry Cao, CFA Application Handbook of Artificial Intelligence and Big Data in Investment from CFA Institute Research Foundation.
If you like this post don’t forget to subscribe enterprising investor.
All posts are the opinion of the author. As such, they should not be construed as investment advice, and the opinions expressed do not necessarily reflect those of CFA Institute or the author’s employer.
Image credit: ©Getty Images / BlackJack3D
Professional Learning for CFA Institute Members
Members of CFA Institute are empowered to self-determine and self-report the professional study (PL) credits earned, including: enterprising investorMembers can easily record credits using credits. Online PL Tracker.