I’ve been pondering how many journalists and podcasters I’ve ticked off when answering this question: Where will AI be in a year?
My answer has always been simple and honest: No one knows.
This week the Chinese start-up DeepSeek is the latest shift in AI’s tectonic plates. Seemingly out of nowhere, it became the top-rated AI app in the United States.
Deepseek’s new, open-source AI model rivals the abilities of OpenAI's most recent model—with far less investment and using reduced-capacity chips. Training costs for ChatGPT-4 were over $100 million. Deepseek’s investment was supposedly $6 million. Deepseek’s model improves energy efficiency making AI more accessible to more people and industries and could be welcomed news when it comes to AI’s environmental impact.
There’s a lot to be learned about Deepseek’s capabilities. For now, it certainly ups the ante in the AI “space race” for global tech titans and governments.
I can hear the hype machines starting to churn as I write this. Here’s a solid piece from Scientific American that outlines what is known.
And so: Where will AI be in a year? No one knows.
T.