BN
|
TechAI Desk•2 views

OpenAI slams Anthropic in memo to shareholders as its leading AI rival gains momentum

OpenAI đã gửi một bản ghi nhớ gửi nhà đầu tư, chỉ trích đối thủ chính Anthropic về chiến lược tính toán (compute) được cho là hạn chế. Theo tài liệu này, OpenAI tuyên bố kế hoạch đạt 30 gigawatt tính toán vào năm 2030, trong khi ước tính Anthropic chỉ đạt khoảng 7 đến 8 gigawatt vào cuối năm 2027. Sự cạnh tranh giữa hai gã khổng lồ AI này ngày càng gay gắt khi cả hai đều chuẩn bị cho các đợt IPO tiềm năng, cùng nhau định vị vị thế trong thị trường mô hình ngôn ngữ lớn (LLM). OpenAI nhấn mạnh lợi thế của mình nhờ vào sự tăng trưởng cơ sở hạ tầng theo cấp số nhân và sản phẩm vượt trội. Trong khi đó, Anthropic, được thành lập bởi các nhân sự cũ của OpenAI, gần đây đã công bố mô hình mới cho sáng kiến an ninh mạng Project Glasswing.

Ad slot
OpenAI slams Anthropic in memo to shareholders as its leading AI rival gains momentum

Open AI CEO Sam Altman speaks during a talk session with SoftBank Group CEO Masayoshi Son at an event titled "Transforming Business through AI" in Tokyo, on Feb. 3, 2025.

With Anthropic gaining momentum in the booming AI market, OpenAI sent a memo to investors this week slamming its chief rival for "operating on a meaningfully smaller curve," and characterizing the company as compute constrained.

OpenAI said it's planning to have 30 gigawatts of compute by 2030, while it expects Anthropic to have roughly 7 to 8 gigawatts by the end of 2027, according to the memo, which was viewed by CNBC.

"Even at the high end of that range, our ramp is materially ahead and widening," OpenAI said. Bloomberg was first to report on the memo.

OpenAI's concern with Anthropic is building as the companies battle for positioning in the rapidly evolving market for large language models and artificial intelligence services. The two companies are collectively valued at over $1 trillion and both are gearing up for IPOs, potentially this year, while trying to convince investors that they have sustainable business models that can withstand the pressure of going up against cash-rich companies like Google and Meta .

Anthropic was founded in 2021 by CEO Dario Amodei and a group of researchers and executives who defected from OpenAI. Earlier this week, Anthropic, which has been winning in the enterprise market, announced a powerful new model that will roll out to a select group of companies as part of a new cybersecurity initiative called Project Glasswing.

Ad slot

In the investor memo, OpenAI pointed to Amodei's comments about his deliberately conservative compute strategy, writing that its own approach is "different."

"Each new generation of infrastructure lets us train more capable models, making every token more intelligent than the one before," OpenAI wrote. "At the same time, algorithmic gains and hardware improvements reduce the cost to serve each token, lowering the cost per unit of intelligence."

OpenAI, which kicked off the generative AI blitz with the launch of ChatGPT in late 2022, said it benefits from "compounding advantage," with better infrastructure and models lowering costs, and superior products leading to higher revenue.

"This leverage also enables OpenAI to continue democratizing AI by making our tools available to hundreds of millions of people for free and being more generous with builders, passing that capacity on to the people who are creating and solving problems with our tools," the company said.

Anthropic didn't immediately respond to a request for comment.

WATCH: OpenAI is chasing Anthropic right now, says Big Technology’s Alex Kantrowitz

Ad slot