Microsoft’s smaller AI model beats the big guys: Meet Phi-4, the efficiency king
Microsoft’s new AI model, Phi-4, outperforms larger competitors like Google’s Gemini Pro with superior mathematical reasoning while using fewer resources.Read More
Read more at VentureBeat
Topics
-
Harvard Is Releasing a Massive Free AI Training Dataset Funded by OpenAI and Microsoft
The project’s leader says that allowing everyone to access the collection of public-domain books will help “level the playing field” in the AI industry.Wired - 1d -
For models of efficiency, Musk and DOGE should ignore DC and look to the states
Chris Sununu, the Governor of New Hampshire, has urged Elon Musk to look to the states for inspiration in streamlining government, citing the success of states with balanced budgets and limited ...The Hill - 1d -
Here’s how big a hit Microsoft is taking on its investment in GM’s Cruise
Microsoft warned investors it will book an impairment charge in its fiscal second quarter related to its investment in GM’s Cruise robotaxi unit back in 2021.MarketWatch - 2d -
Meta launches open source Llama 3.3, shrinking powerful bigger model into smaller size
The 70B-Llama 3.3 is specifically optimized for cost-effective inference, with token generation costs as low as $0.01 per million tokens.VentureBeat - Dec. 6 -
With 4 Words, Google’s CEO Just Fired the Company’s Biggest Shot Yet at Microsoft Over AI
Sundar Pichai says Google’s AI models are competitive, but it’s the product that is the problem.Inc. - Dec. 5 -
Foegele, Kings rally to beat Stars 3-2 for fourth straight win
Warren Foegele scored on a wraparound early in the third period, and the Los Angeles Kings rallied to beat the Dallas Stars 3-2ABC News - Dec. 5 -
Google CEO Sundar Pichai says he’d love to do side-by-side comparison with Microsoft’s AI models
Google CEO fired back at Microsoft CEO's comments aimed at Google's footing in the AI arms race.CNBC - Dec. 4 -
The Kings make 15 straight shots to beat the Rockets 120-111
Domantas Sabonis scored 27 points and Sacramento made 15 straight shots from the field to beat Houston 120-111 and deny the Rockets the top seed in the NBA Cup quarterfinalsABC News - Dec. 4 -
Liquid AI’s new STAR model architecture outshines Transformer efficiency
The STAR framework leverages evolutionary algorithms and a numerical encoding system to balance quality and efficiency in AI models.VentureBeat - Dec. 3
More from VentureBeat
-
Cohere’s smallest, fastest R-series model excels at RAG, reasoning in 23 languages
VentureBeat - 4h -
Esports World Cup Foundation offers $20M in partner club expansion
VentureBeat - 5h -
OpenAI launches ChatGPT Projects, letting you organize files, chats in groups
VentureBeat - 7h -
Shutterstock pioneers ‘research license’ model with Lightricks, lowering barriers to AI training data
VentureBeat - 9h -
Pika 2.0 launches in wake of Sora, integrating your own characters, objects, scenes in new AI videos
VentureBeat - 10h