In summary, Microsoft’s somewhat dispensable product strikes at the core of Google's revenue source.
I started using ChatGPT last year and found it astonishing. Unfortunately, the statistical questions I asked was too profound and after repeated inquiries, I never received a complete response.
Moving on to recent developments, following the rise of ChatGPT (and not just its free public trials), several major companies either launched services linked to ChatGPT or released their versions. This is led by Baidu in China and Google internationally. However, it is not an exaggeration to say that, given the nature of AI text communication, Google and Baidu are unlikely to compete with ChatGPT. This is not due to a lack of technology, but the clear demonstration of first-mover advantage, as evidenced by Google's scrambling efforts to keep up.
After the popularity of ChatGPT, Google introduced its Google-ChatGPT, Bard (indeed, once a useful product becomes popular, the followers do not even deserve their own names). Bard is currently not publicly tested, only providing a few Q&A examples. One of them includes a factual error, claiming that the James Webb Telescope was the first to capture images of exoplanets, which is laughable given this was already achieved in 2005! Mistakes in AI are not surprising, but making a mistake at this juncture gives consumers the impression of Google being ill-prepared for the battle, leaving users to naturally choose the more familiar ChatGPT. If Bard doesn’t want to be labelled as the Google version of ChatGPT, it must outperform Microsoft, which is why such an error becomes critical: it quenches the expectations of investors.
A side note: the first letter in FAANG stands for Meta, previously known as Facebook. A year and a half ago, I was saying that Facebook's market value of $700 billion should definitely exceed a trillion; it was definitely undervalued. Well, I now take back those words; the capital market is pretty accurate. I believe the primary reason for Mark Zuckerberg changing the name to Meta was nothing else but to draw sufficient attention, allowing him to obtain more funding in the hot topic of the Metaverse to avoid bankruptcy. Regrettably, they chose to do this amidst the fervor of the Federal Reserve's quantitative easing (they changed the name in November 2021). As a result, the stock price, already inflated, plummeted from $380 to $90 last year. Why do I say that if Google were to die, it would certainly follow Meta's demise? Because 97% of Meta's income comes from advertising... The only remaining top product, I believe, is Instagram; other sources of income can't compare. However, even Instagram, strong as it is, is experiencing a continuous decline in monthly active users, and investors' expectations are also consistently dropping. I knew about this 97% figure three years ago, and at that time I thought it was great to rely on advertising – it's light on assets, flexible, and responsive to market changes, all of which are advantages. But these advantages can only create cash flow when the economy is on the rise. When Microsoft's ChatGPT + Bing arrives, will Meta's advertising rates take a hit (of course, if forced to discount, Google will not be spared)? (Hopefully, before Facebook's actual demise, it sells off Torch and React!)
Therefore, Microsoft's ChatGPT + Bing is indeed touching Google's bottom line. It's a matter of life and death, which is why they hastily released a beta version that isn't ready for public testing, aiming to stimulate investors' imaginations about the product's future potential. Speaking of this, I can't help but address a question: why do so many publicly traded companies engage in activities unrelated to their main business? For instance, DiDi venturing into autonomous driving last summer (when a bunch of Chinese concept stocks plummeted due to the Chinese concept delisting turmoil + extremely poor performance). All these actions, frankly speaking, are ways for listed companies to create an imaginative space for investors, to tell stories for financing, and to boost stock prices. A company that does everything well and a company that doesn't perform but has great potential and mainstream arrangements will surely attract faster financing in the short term, even if the former's profits are much higher than the latter's. The hurried product releases by companies like Baidu are no different from DiDi last year. Perhaps they never intended to take it seriously because it would burn money! In these times, when everyone is in a state of self-preservation, even banks are half-dead from waves of early loan repayments, and the internet industry is laying off employees, I don't believe they genuinely intend to develop their AI robots.
I'd like to add a note: why is early repayment of a mortgage fatal for banks? Apart from the well-known decrease in bank loan income and accumulation of a large amount of cash with no place to invest, I think a more realistic factor is about the bad debt ratio (bad debt amount divided by total loan amount). So, if the total loan volume keeps decreasing substantially, this percentage would quickly exceed the threshold and trigger a central bank warning.
Google was once a pioneer. Its Chrome browser is far more complex than a computer system, but there hasn't been anything good in nearly 15 years. It missed many opportunities, from the entrance of machine learning, deep learning, and AI into public view around 2013, to cloud computing around 2016, and even the field of virtual currency (such as exchanges, look at FTX, which was only established in 2019). None of these were grasped. Perhaps one of the more impressive aspects was the transformer and self-attention model in 2018, but it seems that this model was quickly caught up with.
When analyzing companies, including analyzing anything, I think it's crucial to focus on the most important aspects. For Microsoft, Bing is merely an added value to its most important products (system + license), optional and dispensable. For Apple, voice recognition, foldable screens, and other such features are also optional; after all, they sell hardware licensed by the Apple system. For Amazon, retail is its core business; everything else, including cloud services, is optional in life-or-death situations (online services and third-party seller services account for over 65% in any quarter so far). But for Google, no matter how well it does with YouTube, Gmail, etc., losing the significant advantage of its search engine in terms of revenue sources would lead to bankruptcy.
Perhaps let's talk about one more insight. In scenarios similar to a self-reinforcing model, wherever there's a game of strategy, such dynamics inevitably exist. For instance, publicly listed companies often embellish or outright falsify their narratives to attract investors. If a subpar company can garner funds based solely on a flashy presentation and happens to have a genuinely ambitious CEO, with the backing of these resources, it might indeed outperform another company that's technologically superior but lacks financial prowess.
Another frequently discussed concept in economics is "anticipated inflation." It suggests that if the public expects a rise in future inflation, then the future inflation rate will indeed increase. This is precisely why Jerome Powell, along with a vast majority of investors, knows he's not being truthful. How could he confidently claim in early 2022 that a 5.5% inflation rate is completely under control and that the future is bright? The reason is that he has to make such claims—and do so with utmost confidence—otherwise, the future inflation rate would be even higher!
This expectation theory also applies to a company's stock price. Unfortunately, Google has woven a faulty story this time.