đź’¨ Abstract
Meta's VP of generative AI, Ahmad Al-Dahle, denied a rumor that the company manipulated benchmarks to make its new AI models, Llama 4 Maverick and Llama 4 Scout, appear more capable than they are. The rumor arose due to reports of poor performance on certain tasks and Meta's use of an experimental version of Maverick for better benchmark scores.
Courtesy: techcrunch.com
Summarized by Einstein Beta 🤖
Suggested
Google is allegedly paying some AI staff to do nothing for a year rather than join rivals
Microsoft reportedly fires staff whose protest interrupted its Copilot event
Amazon says its AI video model can now generate minutes-long clips
How one tweet wreaked havoc on the stock market
Flexport CEO Ryan Petersen’s high-stakes test amid tariff turmoil: 'You can’t be freaking out'
Nikola founder Trevor Milton wants to buy the bankrupt startup's assets
Analyst says Apple, Tesla have biggest exposure to Trump’s tariffs
Former Tesla exec Drew Baglino’s new startup is rethinking the electrical transformer
Waymo may use interior camera data to train generative AI models, sell ads
Powered by MessengerX.io