Silicon Valley Companies Raise Funds For Expensive AI Game
Enterprise AI companies are raising large sums of money to hire engineers to manually make their 'autonomous' products work, according to Crunchbase News. This business model may not be sustainable in the long term.
Original reporting · news.crunchbase.com
Original: “The FDE Bubble: Silicon Valley’s Most Expensive Game Of Pretend”
Read “The FDE Bubble: Silicon Valley’s Most Expensive Game Of Pretend” at Crunchbase News →
How this was written. Hustle Bucks wrote this article with a language model from Crunchbase News’s reporting (model: @cf/meta/llama-3.3-70b-instruct-fp8-fast); it describes that reporting rather than adding to it. The original, linked above, is the full story.
Key points from Crunchbase News’s report
- Enterprise AI companies are raising large sums of money to hire engineers to manually make their 'autonomous' products work
- The business model of these companies relies on hiring large numbers of expensive engineers to make the products work, which may not be sustainable in the long term
- Job postings for 'forward deployed engineer' are up significantly year over year, with high total compensation
- The key question to ask is what happens if the embedded engineer is pulled out tomorrow - does the customer still get what they paid for?
- Software ARR should get cheaper to deliver as it scales, but human-powered ARR does not
Crunchbase News reported that some enterprise AI companies are raising nine-figure rounds to solve a problem that software is supposed to solve by itself: getting humans to use the product. These companies claim to have 'autonomous' and 'agentic' AI platforms, but in reality, they are hiring armies of engineers to manually make the products work.
The business model of these companies involves selling an 'AI platform' to customers, only to discover that the platform does not actually work as promised. To fix this, the companies send in 'forward deployed engineers' to hand-build the workflow that the website implied already existed. The customer is then billed for 'software', and the company books it as annual recurring revenue (ARR).
This model is not a flywheel, but rather a payroll with a pitch deck, according to Crunchbase News. The industry has stopped pretending otherwise, and is now fundraising on it loudly. For example, OpenAI pledged up to $1.5 billion into a deployment joint venture valued at $10 billion, with another $4 billion from other investors.
Job postings for 'forward deployed engineer' are up significantly year over year, with some reports showing an increase of over 1,000%. The total compensation for these engineers is also high, pushing past $300,000 in the US. This has led some to question the sustainability of this business model, as it relies on hiring large numbers of expensive engineers to make the products work.
The key question to ask is what happens if the embedded engineer is pulled out tomorrow - does the customer still get what they paid for? If the answer is no, then the company does not have product-market fit, but rather people-market fit. This has a brutally unglamorous ceiling, as the company can only hire humans so fast, and can never hire them cheap enough to earn a software multiple.
The bill for this business model will eventually come due, according to Crunchbase News. Companies funding their 'forward deployed engineer' armies with fresh nine-figure rounds are betting that nobody will ask the unfashionable questions until after the round closes. However, eventually someone will read the fine print, and the distinction between software ARR and human-powered ARR will become clear.
Software ARR should get cheaper to deliver as it scales, but human-powered ARR does not. This means that companies with this business model may struggle to maintain their growth and profitability in the long term. As the market starts to read the fine print, these companies may find it difficult to sustain their valuations and growth rates.
Questions and answers
What is the business model of these enterprise AI companies?
The business model involves selling an 'AI platform' to customers, only to discover that the platform does not actually work as promised, and then sending in 'forward deployed engineers' to hand-build the workflow.
Why is this business model potentially unsustainable?
The business model relies on hiring large numbers of expensive engineers to make the products work, which may not be sustainable in the long term as the company can only hire humans so fast, and can never hire them cheap enough to earn a software multiple.
What is the key question to ask about these companies?
The key question to ask is what happens if the embedded engineer is pulled out tomorrow - does the customer still get what they paid for? If the answer is no, then the company does not have product-market fit, but rather people-market fit.
