
Exercise 1
Vocabulary
Repeat each word, its meaning and the example sentence out loud.
inference
noun/ˈɪnfərəns/the process of using a trained AI model to produce results from new data
We use inference services to run our model without building our own servers.
annualized revenue
noun phrase/ˌænuəlaɪzd ˈrɛvənuː/money earned in one year, calculated from shorter time periods
The startup's annualized revenue increased to $5 million based on its monthly growth.
compute
noun/kəmˈpjuːt/computer power and processing capacity needed to run software or models
The cost of compute is our biggest expense when running AI models.
gross margins
noun phrase/ɡroʊs ˈmɑrdʒɪnz/the money left after paying for direct costs, shown as a percentage of sales
Our gross margins are falling because we pay more for servers than we did last year.
valuation
noun/ˌvæljuˈeɪʃən/an estimate of how much a company is worth
The company's valuation jumped to $1 billion after its latest funding round.
Exercise 2
Article
Read the article aloud on your own, or listen to it first.
Fal Doubles in Value as AI Inference Demand Grows
Fal, a San Francisco startup, has become one of the fastest-growing companies in AI infrastructure. The company raised $1.3 billion in its latest funding round led by Bessemer Venture Partners.
Other major investors, including the chip maker Nvidia, also contributed to this round. The result is impressive: Fal's value has doubled from $8 billion in April to $16 billion today.
What does Fal do? The company hosts and runs AI models for video, images, and audio.
Specifically, Fal provides inference services. Inference is the process of using an AI model that has already been trained to create new outputs.
Think of it like this: one team builds and trains the AI model, but Fal provides the computer power to actually use that model. Fal's customers can run their AI models through Fal's service without building their own expensive servers.
Fal announced two funding rounds at the same time: Series E and Series F. Together, these rounds brought in $1.65 billion. The earlier round, Series E, was led by Sequoia, a famous investment company.
This rapid funding shows that investors believe in Fal's future and in the growing demand for inference services.
The numbers behind Fal's success are strong. Earlier this year, the company made more than $800 million in annualized revenue.
That means if its monthly growth continues at the same rate, it will earn that much money in a full year. This revenue comes from customers who want to run AI models, especially open source models that anyone can use.
However, Fal faces competition. Other companies like Baseten, Fireworks AI, and Modal also provide inference services, and they are growing quickly and raising large funding rounds too.
But these companies share a common problem: their profit margins are very thin. The reason is simple but serious: buying or leasing the computer power needed to run AI models is extremely expensive.
Even though these companies grow fast and raise billions, they may not make much money from each customer.
Fal was founded in 2021 by Burkay Gur and Gorkem Yurtseven. Gur worked at Coinbase and Oracle as a machine learning expert before starting Fal.
Yurtseven was a developer at Amazon. Both founders brought valuable experience from major technology companies to their new startup.
The success of Fal tells us something important about the AI industry right now. Many companies are building AI models, but they need places to run them.
Fal fills that need, which is why investors are excited about the company's future. As more businesses use AI, demand for inference services will likely keep growing.
Exercise 3
Discussion
Have a discussion based on the following questions.
- 1.Does your company use AI inference services, or do you think you might in the future? Why or why not?
- 2.Fal's margins are thin because computing costs are high. What other tech businesses do you know where costs stay high no matter how much the company grows?
- 3.Would you invest in Fal if you were an investor? What would make you feel confident or worried about this company?
- 4.Fal has strong competitors like Fireworks AI and Modal. In your experience, how important is it to be first in a market, and how important is it to be better than competitors?
- 5.What skills do you think are most important for someone working at an AI infrastructure company like Fal?
Exercise 4
Further discussion
Have a discussion based on the following questions.
- 1.How is AI infrastructure different from other kinds of cloud services, and why do companies need specialized companies like Fal?
- 2.Open source AI models are becoming more popular. How will this change the business of companies like Fal, and what opportunities or problems might it create?
- 3.If computing costs are very high for inference companies, what are some ways they could make more profit in the future?
Discuss this story out loud with our AI speaking coach.
It asks one question at a time, at your level, and gives you feedback at the end.
Start the discussion Retell itSource
This lesson is our own retelling, for English learners, of a story from TechCrunch.



