AI News

Zhang Yiming on ByteDance AI: Catching Up with OpenAI and Anthropic as Early as 2029 or 2030

Cailian PressUpdated Nov 12 16:04

AI summary

At an all-hands meeting of ByteDance's AI team, Zhang Yiming said ByteDance is capable of competing with OpenAI and Anthropic, but truly catching up with frontier technology will take until 2029 or 2030 at the earliest. He also mentioned that compared with top overseas AI companies, the pre-training dataset scale of ByteDance's Seed series models is still relatively small, which is the most pressing issue at present. Foreign media reported that the ByteDance Seed team is pre-training a model with up to 10 trillion parameters.

Why read this

AI score 57

Zhang Yiming gives ByteDance's timeline for catching up with frontier models, and points out that pre-training data scale is the current main gap.

Topics

Tags

Body · AI translation

Cailian Press Financial News Agency

Open the APP

Zhang Yiming speaks again on ByteDance AI: catching up with OpenAI and Anthropic "as early as 2029 or 2030"

① Zhang Yiming said ByteDance is capable of competing with OpenAI and Anthropic, but truly catching up with frontier technology will take "at the earliest 2029 or 2030"; ② Compared with top overseas AI companies, the pre-training dataset scale of ByteDance's Seed series models is still smaller, which is the issue ByteDance AI most needs to focus on right now.

On November 12, The Paper reporters learned from multiple informed sources that recently ByteDance founder Zhang Yiming once again talked about the AI business in an internal sharing session.

It is understood that at a recent all-hands meeting for the AI team, Zhang Yiming said ByteDance is capable of competing with OpenAI and Anthropic, but truly catching up with frontier technology will take "at the earliest 2029 or 2030." ByteDance's large models should benchmark against the international first tier.

Foreign media reported that the ByteDance Seed team is pre-training a model with a parameter scale of up to 10 trillion, a magnitude comparable to Anthropic's most high-end model.

This is not the first time Zhang Yiming has internally focused on the development path of the AI business. Previously, The Paper reporters learned from ByteDance insiders that Zhang Yiming said at an internal meeting that building models requires adhering to long-termism and delayed gratification, rather than using others' outputs to trade for a momentary leaderboard ranking.

Company Operations Insight

Artificial Intelligence

Cailian Press statement: The content of this article is for reference only and does not constitute investment advice. Investors who act on it do so at their own risk.

Business Cooperation

Popular Unlocks

[Intraday Treasure] The inventory cycle may have passed its worst stage; this industry has obvious low-valuation and high-dividend characteristics, institutional holdings in the sector are at historical lows, and this company's flagship product is an exclusive variety

The inventory cycle may have passed its worst stage; this industry has obvious low-valuation and high-dividend characteristics, institutional holdings in the sector are at historical lows, leading companies have stable cash flow and long product life cycles. This company's flagship product is an exclusive variety and it is steadily advancing scientific research and innovation work, while another company is actively exploring the layout of new products.

[Intraday Treasure] Financing enthusiasm for startups in this track remains undiminished; industrialization of this type of product is entering a window period of accelerated advancement, and this company's product has entered the stage of demonstration application verification in typical scenarios

This type of technology is a key research direction, financing enthusiasm for startups in the track remains undiminished, and the industry's industrialization is entering a window period of accelerated advancement. This company's product has entered the stage of demonstration application verification in typical scenarios.

[Gold Medal Minutes Library · Trending Conference] Robot training is moving toward automated simulation and high-quality data expansion, and deep vision, flexible quality inspection and dexterous hand industrialization are ushering in new changes

① Apple's HALO robot simulation system may drive structural adjustments in data collection tasks; simulation data is expected to replace part of visual collection and expand pre-training supply. This company is a representative domestic enterprise in simulation data collection; ② Industrial vision is extending toward flexible quality inspection that autonomously selects detection positions and observation angles, and AI quality inspection scenarios are expected to accelerate penetration. This company is engaged in industrial optical imaging and visual inspection businesses; ③ Robot data collection volume is expected to multiply several times next year and the year after, and tactile data is entering collection gloves and dexterous hand training. Companies providing force and flexible tactile sensors are expected to benefit from related products.

A 131-year-old time-honored brand finally bids farewell to the A-share market: Cuihua Jewelry faces its last trading day tomorrow, with its share price down to only 0.23 yuan

Cailian Press Lu Qinglian 2026-11-11 21:39

① Cuihua Delisting announced that the company's shares resumed trading on October 23, 2026 and entered the delisting consolidation period, with the expected final trading date being November 12, 2026; ② Failure to disclose its 2025 annual report within the statutory period was the direct trigger for the delisting of this century-old gold shop; ③ The Shenzhen Stock Exchange will decide to terminate the listing and trading of the company's shares.

Source

Cailian Press