Meta has claimed that its new artificial intelligence model is the first open-source system that will rival products from competitors such as OpenAI and Anthropic.
In a blogpost, the company said its new model, with the unwieldy name of Llama 3.1 405B, “is competitive” with others – including those from OpenAI and Anthropic – “across a range of tasks”.
If true, it would mean that for the first time, one of the most powerful AI models in the world is available without an intermediary charging for access – or controlling what its technology is used for.
“Developers can fully customise the models for their needs and applications, train on new datasets, and conduct additional fine-tuning,” Meta said. “This enables the broader developer community and the world to more fully realise the power of generative AI. Developers can fully customise for their applications and run in any environment … all without sharing data with Meta.”
Those who use Llama on Meta’s own apps (where it is available, for now, in the US only) will have extra layers of “safety”, the company claims.
Those, too, are open-source, and the company has no way of forcing others to apply them to their own uses of the model.
“I believe that open source is necessary for a positive AI future,” Mark Zuckerberg, Meta’s co-founder, wrote in a blogpost. “AI has more potential than any other modern technology to increase human productivity, creativity, and quality of life – and to accelerate economic growth while unlocking progress in medical and scientific research. Open source will ensure that more people around the world have access to the benefits and opportunities of AI, that power isn’t concentrated in the hands of a small number of companies, and that the technology can be deployed more evenly and safely across society.”
Zuckerberg conceded that “bad actors may be able to use the intelligence of AI models to fabricate entirely new harms”. But, he argued, “I think it will be better to live in a world where AI is widely deployed so that larger actors can check the power of smaller bad actors.”
So far, Meta has been marking its own homework on the question of its model’s power. The model’s sheer size puts it on a par with the largest systems from competitors. But until third parties are able to run fair tests between Llama 3.1 405B and peers such as GPT-4o, it isn’t guaranteed that the pure size will match the effectiveness of the current leaders in the field.
Currently, only users in the US can easily access the AI model. Meta reportedly avoided an EU launch due to concerns about the bloc’s regulatory framework around AI and data protection, and in the end appears to have decided to stick to a single country for the launch period. But as an open-source system, Llama 3.1 405B will probably be accessible to users outside the EU through other channels soon.