Monthly Newsletter - March 2024 - Baskin Wealth Management

How Wise is AI?

You probably have some questions about AI (artificial intelligence). The seemingly instant emergence of multiple platforms offered by the major technology companies has been the biggest event in computing since the advent of social media companies like Facebook and Instagram. I wish I could give you a lot of good answers, but I can’t. The field is advancing (or at least changing) too rapidly. I can give you some thoughts, and I will do so by presenting some actual “dialogs” with Copilot, the AI engine now freely available on Microsoft.

I asked Copilot “What is the difference between a Google Search and a Copilot response?”. Here is what came back (unedited):

Certainly! Let’s explore the differences between a Google search and a Copilot response:
Traditional Search vs. Copilot:
Traditional Search (Google):
Starting Point: In a traditional search engine like Google, you begin by typing in keywords.
Results Pages: You then browse through multiple pages of search results.
Challenges: You might struggle with finding the right keywords or phrasing your query correctly.
Irrelevant Information: Sometimes, you encounter irrelevant information due to incorrect keywords or paid advertisements.
Copilot Response:
AI-Powered Search: Copilot, on the other hand, is an AI-powered search assistant.
Chat-Based Interaction: You can chat with Copilot just like you would with a friend.
Understanding Intent and Context:
Copilot understands your intent and context.
It handles complex and conversational queries effortlessly.
For example, you can ask multi-faceted questions, and Copilot responds with a single comprehensive answer.

I think that’s a pretty good answer. If nothing else, applications like Copilot make it much faster and easier to find the information you want without sorting through multiple webpages, many of which have paid Google to be near the top of the list.

What most people want to know is, do applications like Copilot just regurgitate information, or do they resemble human thinking – that is, real intelligence. I asked Copilot “Is Copilot good at picking stocks?” Here is the answer, again, unedited:

As an AI language model, I don’t have personal opinions or preferences, but I can provide some general information about stock picking.
Stock picking involves selecting individual stocks or securities with the goal of achieving better returns than the overall market. Here are some points to consider:
Challenges of Stock Picking:
Risk: Picking individual stocks can be risky. The performance of a single company can be volatile due to various factors (market conditions, industry trends, company-specific news, etc.).
Research Intensive: Successful stock picking requires thorough research, understanding financial statements, analyzing industry trends, and keeping up with news.
Market Efficiency: The stock market is generally efficient, meaning that most publicly available information is already reflected in stock prices. It’s challenging to consistently outperform the market.
Approaches to Stock Picking:
Fundamental Analysis: Examining a company’s financial health, growth prospects, competitive advantage, and valuation. This involves looking at metrics like earnings, revenue, debt, and management quality.
Technical Analysis: Analyzing historical price patterns, trading volumes, and technical indicators to predict future price movements.
Quantitative Models: Using mathematical models to identify undervalued or overvalued stocks.
Behavioral Analysis: Considering investor behavior and sentiment.
Diversification:
Diversifying your investments across different asset classes (stocks, bonds, real estate, etc.) can reduce risk. Investing in a broad market index (like an S&P 500 ETF) provides diversification without the need for individual stock picking.
Long-Term vs. Short-Term:
Long-term investors often benefit from holding a diversified portfolio rather than trying to time the market or pick individual winners.
Short-term trading (day trading or swing trading) is riskier and requires more active monitoring.
Professional Advice:
Many investors seek advice from financial advisors or use robo-advisors for automated portfolio management.
Some people prefer passive investing (index funds) over active stock picking.
Remember that no one can consistently predict stock movements with certainty. Even professional fund managers often struggle to consistently beat the market. It’s essential to align your investment strategy with your risk tolerance, goals, and time horizon.
Always do your research, consider your risk tolerance, and consider seeking professional advice if you’re unsure.

This is really unhelpful and does not answer the question. A human would say yes, no, or maybe, or perhaps, it is too soon to tell. In this case, Copilot seems to have defaulted to the information you would find on almost every mutual fund’s home page. Another way of saying this is that Copilot has knowledge, but not wisdom.

For me, at least at the moment, that’s the problem. Copilot and other AI applications know everything, but understand little, at least in the way a human understands things. They are superb at performing some jobs; we are already using Copilot to help us find ways to streamline some of our tedious routine tasks, and I have no doubt we will find more and more uses for it. But when it comes to thinking and making decisions, the areas where we apply the hard-learned lessons from decades of experience, I think we are still some way from deferring to the robots.

The Nobel Prize winning inventor of behavioural economics, Daniel Kahneman, died last week. Kahneman was very skeptical about human reasoning and his work highlights how we are all affected (and afflicted) by cognitive biases we are seemingly helpless to avoid. He calls these biases “noise” which leads to a randomness in our decision making. For this reason, he made the assertion: “Clearly AI is going to win against human intelligence”, at least in various domains. My own view is that those domains will be the ones where facts are hard and unyielding, like physics and math. In areas where huge numbers of somewhat irrational humans are trading with each other on the basis of often illogical and frequently false beliefs, which is a pretty accurate description of capital markets, humans will continue to do better than machines. At least for now.

Chairman

David Baskin

Blog

Cash Isn’t Trash Anymore – David Baskin – March 31, 2024

Media Appearances

David Baskin on BNN’s The Street -What are Nvidia’s earnings going to be 3 years from now when its competitors are out there? – March 6, 2024

David Baskin on BNN’s The Street: Investment dollars are leaving Canada – March 6, 2024

Barry Schwartz on BNN’s The Street -There’s good reason to believe things will turn around for the TSX going forward – March 25, 2024

Long Term Investing with Barry Schwartz & Ernest Wong

Financial Planning Edition – March 4, 2024

Buffet, National Bank and Royal way to value CNQ – March 12, 2024

Tourmaline Oil – It’s a gas, gas, gas – March 19, 2024