In a surprise press release, Meta launched its newest AI, the Llama 2 Long, claiming it significantly outperforms rivals GPT-3.5 Turbo and Claude 2 at certain tasks.
The breakthrough reportedly came when the Llama 2 Long was tasked with eating hay and staring into the middle distance. Meta’s AI performed these tasks with 97% more efficiency than its AI counterparts.
In a statement, a Meta spokesperson said, “Our Llama 2 Long AI showcases an unprecedented capacity for chewing cud and looking vacant. It’s a clear indication that we’re leading the AI field in ways that are both groundbreaking and utterly pointless.”
Meanwhile, OpenAI’s GPT-3.5 Turbo and Claude 2 were reportedly unfazed by their lackluster performance in the hay-eating and vacant-staring tests. A representative of OpenAI stated, “Sure, we didn’t focus on these particular tasks. But we’re confident that our AIs still outperform Llama 2 Long on tasks that matter, like drafting emails, writing poetry, and not being a llama.”
Meta has yet to announce whether Llama 2 Long will be marketed as a pet substitute or an office productivity tool. However, insiders suggest the company plans to offer a ‘spit guard’ accessory for customers working in close proximity to the AI.