Key Takeaways:
- Securing workers with AI skills is becoming notably challenging, as acknowledged by the CEO of an AI startup.
- The CEO shared an anecdote where his attempt to recruit an employee from Meta was hindered by a shortage of GPUs, illustrating the competitive nature of talent acquisition in the AI sector.
- He emphasized the importance of offering “amazing incentives” to entice AI talent, indicating that traditional recruitment methods may not suffice in this highly specialized field.
- These insights were shared during a conversation on the podcast “Invest Like The Best,” shedding light on the ongoing struggle to hire skilled professionals in the AI industry.
Recruiting AI talent appears to be a tough feat for some companies.
Aravind Srinivas, the founder and CEO of Perplexity, shared a revealing encounter highlighting the challenges of recruiting individuals with generative AI expertise. In his attempt to hire a senior researcher from Meta, Srinivas was met with a staggering demand: “Come back to me when you have 10,000 H100 GPUs.” This requirement underscores the significant reliance on Nvidia’s H100 GPUs by tech giants like Meta, OpenAI, and Google for powering and training AI chatbots.
However, fulfilling such a demand would entail substantial costs and time investment, estimated by Srinivas to be in the billions and spanning 5 to 10 years. For Perplexity, which utilizes GPT-4 to drive its question-and-answer engine, the combination of limited funds and a chip shortage has compounded the difficulty of securing the talent necessary to develop a large language model.
“People don’t want to leave because when you don’t have anything when they have peers to work with, and when they already have a great experimentation stack and existing models to bootstrap from, for somebody to leave, it’s a lot of work,” the CEO said. “You have to offer such amazing incentives and immediate availability of compute. And we’re not talking of small compute clusters here.”
The CEO added that even if smaller firms like Perplexity are finally able to get Nvidia’s chips, they’ll continue to fall behind because of AI’s rapid speed of development.
That could make it even harder to secure AI talent in the future.
“By the time you waited and got the money and booked the cluster and got it, the guys working here will have already made the next-generation model,” Srinivas said, referring to AI talent at major tech companies.
“They’re like, ‘Look, the world has changed, I’m already in the next generation,'” he added. “‘I’ll come when the next version of the model is finished training. This time, you come back to me when you have 20,000 H100s.'”
Srinivas and Meta didn’t immediately respond to Business Insider’s request for comment before publication.
There has been a rapid uptick in interest in AI skills like machine learning and data engineering since OpenAI launched ChatGPT in November 2022. Companies like Amazon, Netflix, and Meta have offered salaries as high as $900,000 a year to attract generative AI talent, and non-tech companies across the education, healthcare, and legal sectors have been looking to fill roles with workers who know how to use AI.
While Big Tech companies may employ workers who can create AI models that generate desirable outputs, Srinivas believes that skillset alone isn’t enough to make AI tools useful.
“You have to post-train them and address the long tail of issues you get on serving a product,” the CEO said.
Post-training expertise — like knowing how to reduce a chatbot’s factual inaccuracies — is an important skill that employees from industries like crypto or e-commerce can quickly learn, Srinivas said.
Leaning into that skillset, the CEO said, will help AI companies like Perplexity stand out in a sector dominated by Big Tech.
“You have tremendous advantage to create a lot of value,” he said about post-training skills. “And we are focused on that.”