For investors, Meeker's advice is to 'only invest an amount you are willing to lose.'
Mary Meeker, known as the 'Queen of the Internet,' is back, this time targeting the AI giant OpenAI.
Meeker is the founder and general partner of the venture capital firm Bond and was hailed as the 'Queen of the Internet' for her previous annual Internet Trends reports. Before founding Bond, she was responsible for growth at Kleiner Perkins from 2010 to 2019, investing in Facebook, $Spotify Technology (SPOT.US)$ Ring, and Block (then known as Square), among others.
In this lengthy 340-page report titled 'Trends – Artificial Intelligence (AI),' she accurately predicts $Alphabet-A (GOOGL.US)$ 、 $Apple (AAPL.US)$ The rising legendary Analyst clearly states that the growth rate of artificial intelligence surpasses any technological wave in history.
This report continues Meeker's consistent grand vision, encompassing everything from the invention of printing to the application of Roomba robotic vacuums, attempting to depict a panoramic view for the AI era. The report points out that the development speed of AI is unprecedented, with explosive growth in user numbers, usage, and capital expenditure, profoundly changing the global Internet landscape and working methods.
The pace of AI transformation is "unprecedented".
Meeker uses the term "unprecedented" 51 times in the report to describe the speed of AI development, adoption, investment, and usage.
For example, ChatGPT reached 0.8 billion users within 17 months, and this growth rate exceeds any technology in human history.
ChatGPT is also eating into search market share, with its annual search volume reaching 365 billion times, which is $Alphabet-A (GOOGL.US)$ 5.5 times. However, this unprecedented speed of adoption also signifies an equally unprecedented level of competition.
Technological advantages will translate into global dominance.
Most notably is Meeker's determination of the geopolitical impact of AI.
She bluntly stated that 'AI leadership may determine geopolitical dominance,' suggesting that technological competition has escalated to the level of national strategy. For investors, this means that AI-related investments must consider not only commercial returns but also assess the potential impact of geopolitical risks on asset allocation.
India users support the valuation myth of OpenAI.
Report data shows that India has become the largest source of monthly active users for ChatGPT, with usage even surpassing that in the United States.
This means that OpenAI's valuation, which is as high as hundreds of billions of dollars, is largely thanks to the contributions of Indian users.
This finding is quite ironic—an AI giant headquartered in San Francisco and built on Silicon Valley venture capital has its user growth engine sourced from the Emerging Markets across the ocean. More critically, Indian users' sensitivity to price far exceeds that of American users, providing an excellent opportunity for low-cost competitors.
The harsh reality behind the plummeting costs.
The report reveals that while the cost of training top AI models has surged 2400 times over the past eight years, the cost of AI inference has plummeted by 99.7% within just two years.
This scissors effect is reshaping the game rules of the entire industry.
As hardware costs rapidly decline — $NVIDIA (NVDA.US)$ the energy consumption per token of the 2024 Blackwell GPU is reduced by 105,000 times compared to the 2014 Kepler GPU — lightweight models focused on customized scenarios are beginning to show a lethal competitive advantage. They do not require the huge investments like OpenAI but can provide 'good enough' service in specific scenarios.
The combination of the two has caused the costs of top AI products to decline rapidly since their inception. This is good news for consumers, but it also means that companies hoping to fully leverage technology for business benefits will need more robust financial strength.
Chinese AI companies are rising, but the cash-burning model faces a survival crisis.
Data shows that the expected annualized total revenue of the three leading AI companies in the U.S., OpenAI, xAI, and Anthropic, has reached 12 billion dollars, but they have collectively raised 95 billion dollars in funding.
Mickel bluntly pointed out that OpenAI's valuation seems expensive compared to its revenue.
The report noted that these 'aristocratic' AI companies are facing not only cost pressures but also fundamental challenges to their business models with the rapid rise of Chinese models and open-source alternatives.
For investors, Mickel's advice is to 'only invest what you are willing to lose.' She stated:
'Putting all your eggs in one basket is very risky because everything seems to be going up now and looks invincible—until one day when the situation suddenly reverses.'
Editor/joryn
