- AI companies struggle with profitability despite billion-dollar investments
- Price competition leads to razor-thin margins for model access sales
- Startups explore application-layer products and navigate acquisition pressures from Big Tech
Model Mayhem
The AI industry is facing a paradox: billions spent on training sophisticated models like GPT-4 and Claude, yet profit margins are razor-thin.
Aidan Gomez, CEO of Cohere, recently highlighted this conundrum, describing model access sales as a “zero margin business.” The culprit? Fierce competition leading to price dumping and even free model offerings.
The Application Lifeline
Gomez suggests that the real excitement lies in the application layer. Products like OpenAI’s ChatGPT subscription could be the key to generating meaningful revenue.
For now, AI models are burning through cash faster than they can make it, putting immense pressure on startups in the field.
Big Tech’s Hungry Eyes
As the AI landscape evolves, many startups find themselves acqui-hired by large cloud providers. This trend leaves behind unprofitable business models but preserves valuable technology.
Gomez warns against becoming a subsidiary of cloud providers, emphasizing the importance of maintaining independence in this rapidly changing market.
