DeepSeek flips the script | ITPro
There are decades where nothing happens; and there are weeks where decades happen, so the saying goes, and that is certainly true of the final week of January 2025.
In the weeks leading up to recording this episode, we had planned to talk about Stargate – OpenAI, Oracle, and Softbank’s half-a-trillion pound AI investment. We will, but in a very different context because unless you’ve been under a rock, you’ll know that today’s episode has to be all about DeepSeek.
To say the Chinese large language model has caused an upset in the market would be an understatement. Within days of its release, it wiped more than a trillion dollars off the US stock market, and sent big tech into a public spiral.
But where did DeepSeek come from? And what does it mean for the future of the AI sector?
In this episode Jane and Rory are once again joined by Ross Kelly, ITPro’s news and analysis editor, to explore the profound implications of DeepSeek.
Highlights
“I don’t want to say loser here, because, you know, there’s signs that there’s rallying going on in the markets. But Nvidia dropped around 17 points, losing almost $600 billion in market value. Broadcom as well, equally, down around 17 points.”
“You need to remember what happened after the Sputnik situation, though it was again, a game of one upmanship between the USSR and and the United States and the West. Think we will potentially see that between, you know, us and Chinese based companies in years to come.”
“In making the claim that using relatively underpowered chips, specifically chips that China can can still import not even the frontline flagship chips that Nvidia is using today, in making the claim that using just those DeepSeek has been able to make a model that rivals some of OpenAI’s best models, it scared a lot of investors because it throws into question that whole approach.”
Subscribe
Source link
