Analyze an annual report with generative AI: 10-K

- [Instructor] Now let's take a look at generative AI in action using Apple's 10-K report. If you're not familiar, a 10-K report is an annual filing that publicly traded companies submit to the SEC. You might be eager to jump into the financial statements, but the textual elements like management discussion and analysis, and risks and opportunities can provide deep insights into a company's strategy and market conditions. Let's upload the report into ChatGPT. My first prompt is always to ensure that the LLM can read and understand the document. And you can see how quickly this 121-page document is processed and ready for analysis. Let's go ahead and start our analysis. We can ask a few questions. The first question, let's ask, what were Apple's major sources of revenue growth in fiscal 2024? As ChatGPT instantly responds to this question, think about how long it would take flipping through the document or scrolling through the document to find this information. This is one of the great features of large language models, is they can quickly summarize text and make sense of huge amounts of data. So we see here the key contributors to revenue growth. The growth drivers are higher net sales from advertising, growth in App Store activity, and expansion of cloud services. And we see also that this is cited where the information came from. These citations are important because it lets the user know that the LLM is not hallucinating the results. Further in this response, we see that ChatGPT doesn't just give us the numbers, it actually provides context. Highlighting, for example, that iPhone revenue remained flat year-over-year, and iPad and wearables, home and accessories segments, both declined in revenue, with iPad down 6% and wearables down 7%. This is great contextual information that, typically, you would expect from a human analyst. Let's move on to another question and ask. We talked about risks before, so let's ask, what risks did Apple highlight that could impact its future performance? We see again that ChatGPT references the document. And that they've highlighted macroeconomic and geopolitical risks, as well as operational and supply chain risks. Legal, regulatory, and compliance risks. Cybersecurity and IP. Market and competitive. I like the summary that ChatGPT provided here, where they say, "In sum, Apple's risk profile is shaped by its global scale, heavy reliance on complex supply chains, regulatory exposure, and its position in the fast-changing, highly competitive markets." Very succinct. Let's try one more. This one requires the model to do a little bit more interpretation. We're going to ask, how is Apple adapting its product strategy with the introduction of Apple Intelligence and other AI capabilities? We see here the introduction of Apple Intelligence, what came from the 10-K report, and the strategic implications, which are interpretations of the data that was in the report. And we also see R&D and competitive pressure, and a nice conclusion from ChatGPT. Understanding these textual elements can provide a comprehensive view of Apple's business landscape. But remember, this is just the tip of the iceberg. For those interested in a deeper analysis, you could also explore areas like the letter from the CEO, any kind of competitive analysis and risk disclosures. While a thorough analysis will still require reading the entire 10-K, generative AI presents a great way to interact with the report. This is far more useful than simply searching a document for keywords. With generative AI, you can interact with the document as if you were having a conversation with the management team who prepared it. And tools like this can offer invaluable insights into a company's strategy, risks, and market conditions beyond what numerical data alone can provide. And I know this is a lot of information to take in one video. I mean this, after all, like I said, it's 121-page annual report. So to really understand the power of LLMs, I encourage you to try it out on your own.