AI startup funding in China is also a fraction of the US’ — 11% in Q3’24 — which is not surprising given the Chinese gov has throttled their private sector since 2020.
But as the US has charged ahead in the generative AI race, a new class of LLM startups has emerged in China in response.
These “AI tigers,” many of which are backed by China’s big tech companies, are looking to be the country’s answer to the US’ OpenAI and Anthropic.
In July, OpenAI cut off access to its products in China, fueling the fire.
Most recently, Moonshot AI raised $300M in August (following a $1B round in February), and Baichuan AI raised nearly $700M in July.
Like in the US, China’s big tech has overlapping alliances in the space, with cloud commitments also factoring into deals. E-commerce giant Alibaba has invested in all 5 startups, and tech conglomerate Tencent in 4.
HongShan — Sequoia Capital’s China investment arm that it announced it would spin off in June 2023 — has backed 3 of the companies.
Outside of the US, Saudi Aramco’s Prosperity7 Ventures fund is the most notable foreign investor in the group. The fund invested in Zhipu AI’s Series C in May 2024.
Looking ahead
While the US still leads in AI development, China is closing the gap. In recent months, model performance, especially among open-source models, has started approaching that of US rivals.
Open competition in the US will likely lead to more innovation and big players. But China's heavy state backing could be a competitive edge for developing intensely resource-hungry models — Baichuan AI, for example, has funding from Beijing, Shanghai, and Shenzhen governments.
It will be worth watching China’s AI ecosystem for the eventual AI applications that take root — especially those that become “killer” consumer apps.
MiniMax, 01 AI, and ByteDance (behind the wildly popular TikTok) have all released AI apps in the US, including avatar chatbots, photo-editing, productivity, and search apps.
TLDR — Tech loves drama, right?
Here’s a roundup of our favorite tech drama from last week:
Job security: Former FTX executive Ryan Salame updated his LinkedIn: “I’m happy to share that I’m starting a new position as Inmate at FCI Cumberland!” He is starting a 7.5 year fraud sentence. People on LI congratulated him reflexively as LI users do.
Perplexed: The New York Times sent a cease and desist letter to Perplexity to get the AI search startup to stop using its content. The publisher also has a pending lawsuit against OpenAI. Perplexity said it’s interested in working with the NYT and has “no interest in being anyone’s antagonist here.”
Walking out: 3 Amazon execs are leaving its Just Walk Out division — the company’s attempt to use AI to remove the need for cashiers from stores. The tech has hit a few speed bumps. Its competitor Grabango announced it would shut down last week.
Founder mode: A pic of a startup co-founder whipping out his laptop at his own wedding to complete a pull request generated both outrage and reverence online.Pic here.
Dumber than a cat: In an interview with the WSJ, Meta senior researcher Yann LeCun called out the limitations of LLMs. Does AI pose an existential threat? “You’re going to have to pardon my French, but that’s complete B.S.” He tweeted models remain far from being as smart as cats.
Machines of loving grace: Meanwhile, Anthropic CEO Dario Amodei published a 15K word essay (“Machines of Loving Grace”) on AI’s transformative potential, assuming we arrive at a “powerful AI.” Here’s Claude’s 1-liner summary: “AI: Potentially compressing a century of progress into a decade, if we play our cards right.”
Ok Google: The Department of Justice released its initial proposal for remedying Google’s search engine dominance, including breaking up the company. The 32-page filing references both “behavioral and structural remedies.”
Beef: The beef between Jason Calacanis (co-host of the All-In podcast, early Uber investor) and Palmer Luckey (Oculus VR founder, Anduril founder) has been rekindled. Last week, Calacanis pushed Luckey/Zuck to give more info about his exit from Meta (see how that ended below) after Luckey accepted Meta’s CTO’s public apology. Back in 2022, Luckey called out Calacanis for being a hypocrite at the All-In Summit.
