This is on top of the exorbitant amount of water required for cooling per query, with GPT-4 consuming up to four times more water for cooling than previously thought—three water bottles to generate a mere 100 words.

Some AI researchers hailed DeepSeek’s R1 as a breakthrough on the same level as DeepMind’s AlphaZero, a 2017 model that became superhuman at the board games Chess and Go by purely playing against itself and improving, rather than observing any human games.

The Chinese startup DeepSeek released an AI reasoning model that appears to rival the abilities of a frontier model from OpenAI, the maker of ChatGPT.

The announcement confirms one of two rumors that circled the internet this week. The other was about superintelligence.

DeepSeek's friendly whale hearkens back to a more playful era of tech branding—and it might just be the disruptor the AI industry needs.

Hedge fund manager and entrepreneur Liang Wenfeng built an AI model on a tight budget despite US attempts to halt China’s high-tech ambitions.

Users across various platforms have reported instances where OpenAI's o1 model begins its reasoning process in English but unexpectedly shifts to Chinese, Persian, or other languages

Researchers are identifying current and future dangers within AI models away from the conflicts of interest they’d face in the industry

Mistral, the French AI lab, is working toward an initial public offering, co-founder and CEO Arthur Mensch said in an interview at Davos.

R1, sent shockwaves through Wall Street, with major tech firms—most notably Nvidia—experiencing sharp stock declines.

Google's DeepMind leads the market with Gemini, TPU chips, and cloud growth, making its valuations compelling compared to its peers. Click for our GOOG update.

A string of startups are racing to build models that can produce better and better software. They claim it’s the shortest path to AGI.