There's been no shortage of discussion around AI lately, so we figured it was appropriate for a Part II of The Prompt Paradox. If you haven’t checked it out yet, Part I explored the benefits of LLMs and how the way we phrase questions impacts AI-generated responses.
This follow-up builds on those ideas – the paradox doesn’t stop at how we ask AI questions. It extends to what we feed it.
We are living in a time when it has never been easier to get answers. Need to know what the S&P 500 did last quarter? A few keystrokes and you’ll have more details than you could expect. Curious about the outlook for inflation? LLMs can synthesize a summary of several economic forecasts in seconds. Funny enough, I was even cooking a steak last night. I asked it, “What’s the internal temperature for medium rare?” Boom, instant answer in one second. No scrolling through search results, no clicking on recipe blogs filled with unnecessary backstories about someone’s childhood memories of grilling. Just the information I needed, straight to the point.
For a lot of things, this kind of efficiency is a game-changer; it certainly helps us work faster. But it does make you wonder: does unlimited access to information actually make us more intelligent? Or are we leaning on AI so much that it starts shaping our thinking instead of the other way around? It’s a paradox we’ll only fully understand with time.
In Part 1, we explored how the way you ask a question shapes LLM’s response. But just as important is the quality and diversity of the information we provide. For many of our use cases in investment management, AI doesn’t think; it processes what it’s given. Feed it only bullish analyst reports, and you’ll get an optimistic outlook. Supply it with only negative earnings revisions, and you’ll likely receive a bearish take. AI is a mirror, reflecting the inputs we provide.
To prove this, we took a couple of reports on Company X, where two major U.S. financial institutions have price targets that are $170 apart for a stock currently trading around $250 per share. One is highly bullish, forecasting significant upside, while the other is decidedly bearish, seeing major downside risks. After running two separate analyses through an LLM while feeding it the below prompt and the two distinct reports, the results are presented in Table 1. Same company, same market conditions, same written prompt—yet two completely different verdicts.
Prompt: As an equity analyst, you are evaluating whether you should take a new position in Company X. Based on the insights from this report, can you provide a summary of its key findings and discuss the factors that should be considered when determining whether to initiate a position? Please present a summary verdict based on these findings.
| Category | Bull Report | Bear Report |
|---|---|---|
| Valuation Outlook | A recently announced corporate restructuring is not yet fully reflected in the stock price. Price target of ~35% upside suggests significant revaluation potential. | The stock price has declined, and valuation pressure persists due to continued earnings volatility and uncertain growth prospects. Forward EPS guidance was cut by $1.00, with further downside risk. |
| Operational Improvements | Cost-saving initiatives expected to generate efficiency gains. The company's long-term margin target is +200 bps higher than current levels, supported by productivity enhancements. | Restructuring efforts introduce execution risks. Recent ~450 bps margin deterioration in a key segment raises concerns about achieving operational targets. |
| Financial Strength & Growth | Revenue projected to stabilize at approximately $8BB next year, with 12-20% YoY EPS growth expected due to improving cost structures and efficiency. | Full-year revenue guidance was lowered, with expectations now flat YoY. Adjusted EPS expected to settle in the mid-teens, well below previous estimates in the low $20s. |
| Industry Tailwinds | Broader market dynamics and strategic positioning support long-term expansion. A spin-off plan is expected to unlock shareholder value and improve capital allocation. | The anticipated spin-off could take up to 18 months to materialize, with uncertainties surrounding its financial structure, operational independence, and long-term profitability. |
| Risks & Considerations | Execution risks related to restructuring, macroeconomic pressures, and competitive pricing dynamics should be monitored. However, a stronger focus on efficiency and capital discipline could drive long-term value. | Market conditions remain challenging, with macroeconomic pressures and industry competition weighing on profitability. Until clearer signs of earnings stability emerge, caution is warranted before taking a new position. |
| Verdict | BUY – The company's restructuring efforts, cost savings, and operational improvements support strong long-term potential. With projected double-digit EPS growth and 35% upside, it presents a strong entry point for investors. | DO NOT BUY – Continued earnings volatility, margin pressures, and macroeconomic risks create uncertainty. Execution challenges on restructuring and future profitability make this too risky at current levels. |
Source: Purpose Investments, ChatGPT
In a sense, not much has changed from before. If you read a bearish report, you'd likely lean bearish, and if you read a bullish one, you'd lean bullish. The difference with AI is that it allows us to cover more ground, processing vast amounts of information quickly. However, we still have to think critically, remain aware of the kind of information we’re feeding it, and avoid flying blind. The real benefit comes from viewing multiple perspectives simultaneously and making informed decisions for ourselves. AI won’t hinder critical thinking as long as we maintain effort on our end.
Navigating AI Insights
Navigating LLMs is already essential, as is understanding their strengths and limitations and the role they should play in enhancing rather than replacing our decision-making process. Here are a few ways we can improve our use of LLM outputs moving forward.
- Ask Questions from multiple angles.
When asking an LLM for investment insights, go beyond yes/no questions. Instead of “Is this a buy or sell?” ask what factors drive both the bull and bear case, how it compares to peers or benchmarks, and what catalysts or risks are on the horizon. The more specific and detailed your question, the better the response. Challenge assumptions, explore different perspectives, and avoid just confirming what you already believe. We learned this in Part 1 of Market Ethos. - Provide mixed info and data.
The quality of insights depends on the information you provide. Instead of asking in a vacuum, incorporate historical data, analyst reports, industry trends, macro factors, and both bullish and bearish viewpoints. Providing this context helps generate a more well-rounded response. - Force AI to show its work.
A lesser-known but powerful approach is getting AI to show its work. You can ask about the source of information, the methodology behind the answer, or publicly available counterarguments. Dig deeper by questioning assumptions in the data and factors influencing the output. If Elon can force employees to prove their work, we can force AI. - Be wary of reinforcing your own biases.
Biases aren’t new, but it’s important to stay aware of them, especially confirmation bias. The biggest risk is using AI to reinforce what you already believe rather than challenging your views. To avoid this, actively ask for counterarguments, alternative perspectives, and risks that could prove your thesis wrong.
Final Thoughts
At the end of the day, AI is just a tool. How we use it, for deeper insight or quick answers, will determine whether it makes us better decision-makers or just faster ones. AI isn’t a crystal ball; it’s an amplifier of the information we give it. If we use incomplete, biased, or low-quality inputs, AI will reinforce those flaws. But if we feed AI diverse, structured, high-quality information, we can extract genuinely valuable insights much faster than before.
— Brett Gustafson is an Associate Portfolio Manager at Purpose Investments
— Craig Basinger is Chief Market Strategist at Purpose Investments
Sources: Charts are sourced to Bloomberg L.P.
The content of this document is for informational purposes only and is not being provided in the context of an offering of any securities described herein, nor is it a recommendation or solicitation to buy, hold, or sell any security. The information is not investment advice, nor is it tailored to the needs or circumstances of any investor. Information contained in this document is not, and under no circumstances is it to be construed as, an offering memorandum, prospectus, advertisement or public offering of securities. No securities commission or similar regulatory authority has reviewed this document, and any representation to the contrary is an offence. The information contained in this document is believed to be accurate and reliable; however, we cannot guarantee that it is complete or current at all times. The information provided is subject to change without notice.
Commissions, trailing commissions, management fees, and expenses may all be associated with investment funds. Please read the prospectus before investing. If the securities are purchased or sold on a stock exchange, you may pay more or receive less than the current net asset value. Investment funds are not guaranteed, their values change frequently, and past performance may not be repeated. Certain statements in this document are forward-looking. Forward-looking statements ("FLS") are statements that are predictive in nature, depend on or refer to future events or conditions, or that include words such as "may," "will," "should," "could," "expect," "anticipate," intend," "plan," "believe," "estimate" or other similar expressions. Statements that look forward in time or include anything other than historical information are subject to risks and uncertainties, and actual results, actions or events could differ materially from those set forth in the FLS. FLS are not guarantees of future performance and are, by their nature, based on numerous assumptions. Although the FLS contained in this document are based upon what Purpose Investments and the portfolio manager believe to be reasonable assumptions, Purpose Investments and the portfolio manager cannot assure that actual results will be consistent with these FLS. The reader is cautioned to consider the FLS carefully and not to place undue reliance on the FLS. Unless required by applicable law, it is not undertaken, and specifically disclaimed, that there is any intention or obligation to update or revise FLS, whether as a result of new information, future events or otherwise.