Business

American government goes everywhere exporting its brand of democracy. The Chinese government wants to do business and make money [with these autocracies] in Asia, Africa, and South America - Lee Kuan Yew In many ways, American companies take the same approach. They build one product, localize it a bit for the target countries, and try to win big there. However, Chinese companies seem to take a different approach. China utilizes lower labor costs than its American counterparts and builds targeted products for markets. And Temu is the perfect example of that. Temu’s parent company’s flagship product is Pinduoduo. Pinduoduo is designed for the Chinese market. A completely unrelated E-commerce product, Temu is designed for the American market.

Every business has revenues and costs. When the revenues and costs are not aligned, the business sooner or later risks bankruptcy. Let me illustrate with a few examples. A restaurant’s costs consist of raw food materials and labor. While the revenues are a function of the amount of food sold. So, in case the revenue falls, the cost of labor kills the restaurant business. A cloud kitchen, due to its reduced labor costs, is more resilient. The costs of oil (or mining) companies consist of drilling and transportation. While the revenue is a strong function of the unpredictable oil/mineral price. That’s why small oil and mining companies go out of business whenever there is a sharp fall in the price of the commodity sold. An airline’s costs consist primarily of the predictable cost of leasing the airplanes and the unpredictable oil prices to fly the plane. While the revenue consists of the number of seats sold well in advance. So, any fall in seats sold or any sudden spikes in oil prices leads to a disaster. A software company’s costs consist mostly of labor and infrastructure. While the revenue comes from the increased sales of the software. The incremental cost of serving a user is marginal, while the fixed costs are huge. That’s why landing a few big initial contracts to become default-alive can make a huge difference in the long run. An exchange takes a cut of transactions flowing through it. The costs and revenues are aligned. Further, if it is an electronic exchange, like stock exchanges or Ad exchanges, then the infrastructural costs are minimal and employee costs are less material. This becomes a pure money-making exercise.

The true value proposition of LLMs lies in their ability to convert unstructured data from sources like websites and documents into structured information with reasonably high accuracy. Yet, the real profit lies in the products built on top of LLM technology. Each year, approximately 4 million books are published worldwide. On average, a book contains fewer than 120,000 words, translating to less than 160,000 tokens in LLM (Large Language Model) terms. Imagine if every single one of these books were generated by GPT-4—it would amount to an astounding 640 billion tokens. At $5 per million tokens, generating all these books would tally up to about $3.2 million! Let’s say the book market represents only about 1% of the total LLM text generation opportunity. Even then the total addressable market of LLM text generation is approximately $300 million annually—a modest figure when compared to AWS, which raked in $90 billion in 2023 as the cloud market leader.

A few years back, I scraped data on podcasters from iTunes. The data was a bit underwhelming and made me realize that podcasters can’t be a potential market. It is a bit dated but I believe is still relevant.

Two cryptocurrency exchanges came out early on from Y Combinator. One is 2012. One in 2013. One returned 1500X to early investors. The other one ceased to exist after 2 years. What happened?

The Half-Life of Facts is a provocative book that makes us realize that a lot of facts that we believe are factually true get overturned over time. It gives several examples of facts that were later proven wrong or became more precise as our measurements and understanding improved.

The early adopters of Instagram were not VCs.

In the decade of the 2010s, after Facebook became mainstream, social became a building block for websites. A lot of websites added a concept of user profiles consisting of a name, a display picture, a profile description, and the ability to see/follow the activity stream of other users. Slowly, this became the standard expectation. ChatGPT will have a similar impact. More and more users will get used to asking questions in natural language. Eventually, all the FAQ and knowledge-base articles would be replaced with a single box where users will ask a question in their preferred natural language. And then even the idea of actions would switch over to that one box, that users can use to navigate around and perform actions in the web applications.

A charity is when your customers capture value when you don’t. A scam is when you capture value but your customers don’t. A business is when you and your customers both capture value. - unknown

I’m a big of serverless. In fact, in 2021, I believe that it should be the default choice and one should have a strong reason to deploy their own servers.