Why should companies avoid building their own custom AI models from scratch?
Replied byJensen Huang
Co-founder & CEO at NVIDIA
Niche: AI
Revenue: Approx. 18 Billion USD/month
Location: Santa Clara, California, United States
Started: 1993
Building and hosting your own AI is incredibly clumsy, expensive, and time-consuming. Most companies should rent pre-built intelligence from frontier labs like OpenAI or Anthropic because it is more convenient and cost-effective than running it themselves.
0
From the Full Interview
This answer is part of a full interview with Jensen Huang, Co-founder & CEO at NVIDIA.
Share this Answer
Found this insight valuable? Share it with your network to help others learn from Jensen Huang's experience.
Cite This Answer
Use this answer in your research, article, or academic work
Related Answers
How should enterprise leaders structure their budgets between experimentation and production AI?
By Sravanth Aluru
AI
Not Publicly Disclosed/mo
Why, if you believe in these companies so strongly, are their leaders publicly raising existential alarms?
By Jensen Huang
AI
Approx. 18 Billion USD/mo
Governments regulate cars and airlines and medical devices precisely because companies cannot be trusted to do it alone. Why should the most powerful technology in history be different?
By Jensen Huang
AI
Approx. 18 Billion USD/mo
How should developers decompose tasks when building with system one models?
By Diogo Almeida
AI
Pre-revenue/mo
When should you build an agent versus when should you build an app?
By Anton Oika
AI
Approx. USD 45 Million/mo
If someone is starting a company tomorrow, what should their primary goal be?
By Rahul Vohra
AI
Approx. USD 30 Million/mo
Why is it so difficult for companies in the AI industry to slow down their development pace?
By Dario Amodei
AI
Approx. $390 Million/mo