Meta AI
Meta's AI division, formerly FAIR, known for open-source models and research on recommendation systems, NLP, and computer vision.
Think of it like
The company that pushes open-source aggressively, believing good science beats proprietary lock-in.
Example
Meta released Llama (open-weight LLM, 2023), LLaMA 2 (2023), and invested heavily in multilingual models and vision systems.
How it actually works
Meta's strategy is different from OpenAI and Google: release open weights, lead through scale and quality, let the ecosystem build on top. They face pressure to monetize LLMs but have invested in research leadership.
For product teams
Open-source strategy builds goodwill and fast research iteration; harder to monetize directly.
For engineers
Llama's weights are Apache 2.0 licensed; enabled a wave of fine-tuning and distillation startups.
Read anything AI without the jargon
Look up any term in plain English, or save terms as you read with the free Chrome extension.
Open DecoderAdd to Chrome