Meta and Microsoft release Llama 2, an AI language model for commercial use<p>The rumors of a <a data-i13n="cpos:1;pos:1" href="
https://www.engadget.com/meta-could-soon-make-its-ai-model-available-for-commercial-projects-114021749.html">commercially-oriented Meta AI model[/url] were true. Meta and Microsoft have teamed up to <a data-i13n="cpos:2;pos:1" href="
https://about.fb.com/news/2023/07/llama-2/">unveil[/url] Llama 2, a next-generation large language (very generalized) AI model intended for both commercial and research purposes. The upgraded open source code places a greater emphasis on responsibility. Developers "red-teamed" models (that is, tested them for safety) and created a transparency schematic to detail potential issues. They also include a responsible use guide, and there's an acceptable use policy to prevent abuses like criminal activity, misleading representations and spam.</p><p>Meta is <a data-i13n="cpos:3;pos:1" href="
https://ai.meta.com/llama/">releasing[/url] both pre-trained and conversation-oriented versions of Llama 2 for free. Microsoft is making it available through the Azure AI catalog to use with cloud tools, such as content filtering. The tool can also run directly on Windows PCs, and will be available through outside providers like Amazon Web Services and Hugging Face.</p><span id="end-legacy-contents"></span><p>The original AI model was open source, but intended strictly for academics and researchers. Llama 2 lets companies customize the technology for their own purposes, such as chatbots and image generators. Ideally, this lets companies benefit from the modelling while providing outsiders a way to check for biases, inaccuracies and other flaws. Open source code isn't new in the AI world — Stability's Stable Diffusion is a notable example. However, major rivals like OpenAI's GPT-4 tend to be locked down for the sake of subscription or licensing revenue. There are also concerns that hackers and other bad actors may ignore open source licenses and use those tools for dangerous purposes.</p><p>As with GPT-4 and Anthropic's <a data-i13n="cpos:4;pos:1" href="
https://www.engadget.com/anthropic-releases-claude-2-a-more-capable-less-gullible-ai-chatbot-160053168.html">Claude 2[/url], the emphasis on responsible use isn't surprising. The tech industry is concerned that large language AI models <a data-i13n="cpos:5;pos:1" href="
https://www.engadget.com/ai-pioneer-geoffrey-hinton-isnt-convinced-good-ai-will-triumph-over-bad-ai-181536702.html">may get out of control[/url], leading to killer robots or the further spread of misinformation. Experts and company leaders have even <a data-i13n="cpos:6;pos:1" href="
https://www.engadget.com/tech-leaders-and-ai-experts-demand-a-six-month-pause-on-out-of-control-ai-experiments-114553864.html">called for a six-month freeze[/url] on experimentation to ensure developers tackle ethical and safety concerns. Politicians also hope to regulate AI, with one Senate bill intended to <a data-i13n="cpos:7;pos:1" href="
https://www.engadget.com/senate-bill-would-hold-ai-companies-liable-for-harmful-content-212340911.html">hold AI creators liable[/url] for damaging content.</p><p>For Microsoft, Llama 2 represents another bid to stay ahead of AI rivals <a data-i13n="cpos:8;pos:1" href="
https://www.engadget.com/google-unveils-its-multilingual-code-generating-palm-2-language-model-180805304.html">like Google[/url]. Microsoft already makes use of OpenAI systems in products like Azure and <a data-i13n="cpos:9;pos:1" href="
https://www.engadget.com/microsoft-confirms-bing-runs-on-the-new-gpt-4-model-184121316.html">Bing[/url]. The Meta collaboration gives Microsoft's business customers more choices, particularly if they're interested in fine-tuning a model to suit their needs.</p>This article originally appeared on Engadget at
https://www.engadget.com/meta-and-microsoft-release-llama-2-an-ai-language-model-for-commercial-use-163615807.html?src=rssSource:
Meta and Microsoft release Llama 2, an AI language model for commercial use
