IN FOCUS
In today’s newsletter, we look at:
- DeepSeek’s bid for AI supremacy
- Alibaba’s attempt to seize the AI spotlight
Welcome to The Prompt! Delivered every fortnight, this free newsletter breaks down the biggest stories and trends in AI. Get all our content, including The Prompt, by registering for a Tech in Asia account. Alternatively, sign up here if you just want to receive this newsletter.
Hi there,
Recently, a 2023 video clip of OpenAI founder Sam Altman popped up on my Instagram feed.
In it, Altman, while in India, was asked whether a small team with just US$10 million could build something truly significant in AI. The person who asked the question was none other than PeakXV Partners managing director Rajan Anandan.
Altman’s blunt reply: “It’s totally hopeless to compete with us.”
A year and a half later, China-based AI firm DeepSeek claims to have produced a model that performs as well as OpenAI’s GPT-4, which the startup developed using only US$6 million. To put that in perspective, GPT-4 is estimated to have cost between US$41 million and US$78 million to train.
US tech stocks lost nearly US$1 trillion as a result, although they have since recovered some of those losses. This is a lesson for all founders not to underestimate competition, even – or maybe especially – when it comes from those who don’t have as many resources.
In the Big Story this week, my colleague Scott analyzes how DeepSeek kept its costs down.
Meanwhile, in the Hot Take, I look at Qwen 2.5-Max, the latest AI release from another Chinese player, Alibaba.
One thing is clear, 2025 will be a year of major developments in genAI. Let’s just hope it’s not one of mishaps.
— Samreen
THE BIG STORY
Chips don’t lie: Did China’s latest AI just beat the US?
THE HOT TAKE
DEEP READS
Stay updated on the go with our mobile app.
Get latest insights with smoother, more personalized experience through TIA mobile app.
