News

Anjney Midha, a16z VC and Mistral board member claims DeepSeek won’t quell AI’s GPU need.

At that time, DeepSeek unveiled Coder V2, which, according to a report it published last year, competed with OpenAI’s GPT4-Turbo for jobs involving coding. According to him, this set DeepSeek on a course to deliver updated models every few months through R1. Its latest open-source reasoning model, R1, has revolutionized the tech sector by providing industry-standard performance at a significantly lower cost.

According to Midha, R1 does not imply that AI fundamental models would cease investing billions of dollars to consume GPU chips and build up more data centers.

It means they will do more with the computing power they can obtain.

“When people are like, okay Anj, Mistral has raised a billion dollars,” he says. “Does DeepSeek mean that all that billion dollars is completely unnecessary? No, actually, it’s extraordinarily valuable for them to be able to look at DeepSeek’s efficiency improvements, internalize them, and then throw a billion dollars at it.”

“Now we can get 10 times more output from the same compute.”

He contends that this does not imply that Mistral is utterly behind competitors Anthropic and OpenAI. Compared to Mistral, each of them has raised billions more. According to reports, OpenAI is in negotiations to raise an additional astounding $40 billion.

He claims that because Mistral is open source, it is still competitive with them. And there is some truth to his reasoning. Because they use the project, open source allows a business to receive practically free technical labor from volunteers. Closed-source competitors must pay for all labor and processing power and protect their trade secrets.

“You don’t need $20 billion. You just need more compute than any other open source model app. So Mistral is positioned [well]. They have the most compute of any open source provider,” Midha said of his portfolio company.

The biggest open-source AI model competitor to Mistral in the West, Facebook’s Llama, will also receive a significant increase in funding. On Wednesday, CEO Mark Zuckerberg stated that he still intends to invest “hundreds of billions of dollars” in AI. This includes $60 billion for capital projects, primarily data centers, in 2025.

Read Also:

  1. How To Add Music To Instagram Post?
  2. NASA Perseverance Rover Finds Strange-Looking Zebra-Striped Rock on Mars
  3. How To Play Google Tic-Tac-Toe?

News Source

Avijit Sah

Avijit Sah is a digital marketing expert specializing in SEO, social media, and content strategy. With a passion for helping businesses grow online, Avijit Sah uses data-driven tactics to boost visibility and engagement. Follow Avijit for the latest digital marketing tips and insights.

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button