All articles tagged with #open source ai
Originally Published 5 months ago — by Fortune
Researchers found that filtering risky content from AI training data, such as bioweapons instructions, can create safer models without harming performance, highlighting the importance of pre-training safeguards over post-training tweaks. The study emphasizes transparency and proactive safety measures in AI development, contrasting with industry secrecy.
Originally Published 1 year ago — by Meta Store
Meta's open source Llama AI models are being utilized by U.S. government agencies and private sector partners to enhance national security and defense applications. Collaborations with companies like Oracle, Lockheed Martin, and Microsoft aim to streamline operations and improve technological capabilities. This initiative supports U.S. leadership in AI, promoting ethical and responsible use while fostering global standards. The widespread adoption of open source AI is seen as crucial for economic and security interests, with the U.S. aiming to set global standards in AI development.
Originally Published 1 year ago — by The New York Times
Mark Zuckerberg has regained popularity in Silicon Valley by championing open-source artificial intelligence through Meta's AI system, which allows developers to freely access, modify, and reuse the code, contrasting with the closed models of companies like Google, OpenAI, and Microsoft.
Originally Published 1 year ago — by VentureBeat
Abacus AI has released Smaug-72B, a new open-source language model that outperforms GPT-3.5 and Mistral Medium, marking a significant milestone in the field of natural language processing. With an average score of over 80 on the Hugging Face Open LLM leaderboard, Smaug-72B showcases the potential of open-source AI to rival Big Tech's capabilities. This release, along with Qwen's recent advancements, signals a new era of AI innovation and democratization, challenging the dominance of big tech companies and opening new possibilities for everyone.
Originally Published 2 years ago — by Ars Technica
IBM and Meta have formed the AI Alliance, a coalition of over 50 organizations including AMD, Intel, NASA, CERN, and Harvard University, with the goal of promoting open innovation and open science in AI. The alliance aims to provide alternatives to closed AI systems used by market leaders like OpenAI and Google, and encourages open AI research and the sharing of AI model weights and research. The AI Alliance plans to develop benchmarks, support AI hardware accelerators, and promote diversity in AI foundation models.
Originally Published 2 years ago — by The Associated Press
Tech giants are divided over the future of artificial intelligence (AI), with Facebook Meta and IBM advocating for an "open science" approach to AI development, while Google, Microsoft, and OpenAI prefer a closed approach. The open camp believes in making AI widely accessible, while the closed camp prioritizes safety and profit. The debate revolves around whether AI should be built in a proprietary and closed manner or with open-source principles. OpenAI, despite its name, builds closed AI systems due to commercial incentives and concerns about the potential dangers of powerful AI falling into the wrong hands. Critics argue that open-source AI models pose risks, such as disinformation campaigns, while proponents argue for the benefits of open-source AI in reflecting human knowledge and culture. Governments, including the US and the EU, are grappling with how to regulate AI and manage the potential benefits and risks of open-source AI components.
Originally Published 2 years ago — by CNBC
The CEO of Google DeepMind, Demis Hassabis, has refuted claims made by Meta's chief AI scientist, Yann LeCun, that DeepMind is engaging in corporate lobbying to control the regulation of AI. Hassabis stated that DeepMind is not seeking "regulatory capture" and emphasized the importance of addressing near-term harms, misuse of AI by bad actors, and long-term risks associated with artificial general intelligence. He also expressed the need for international cooperation and dialogue on AI regulation, including engagement with China. LeCun advocates for open-source AI and warns against a regulatory approach that could stifle open-source models and concentrate AI control in the hands of a few companies.
Originally Published 2 years ago — by Business Insider
Marc Andreessen believes that AI can "make everything we care about better" and that AI companies should be able to build fast and aggressively without regulation, in order to maximize its gains for economic productivity and human potential. He disagrees with the idea of regulating AI and believes that the future of AI should be decided by the free market. He argues that open source AI should be allowed to spread freely and compete with commercial AI companies and startups. To offset the risks of AI being used for nefarious purposes and to block China from becoming an AI superpower, the private sector needs to work with governments to come up with solutions.
Originally Published 2 years ago — by VentureBeat
Emad Mostaque, the CEO of unicorn startup Stability AI, has responded to a recent Forbes investigation that revealed he lied about his background, achievements, and partnerships. Mostaque claims that Forbes ignored factual corrections and made false allegations. He also downplayed his co-founders' and employees' role in the success of Stable Diffusion, an open-source AI text-to-image generator. Despite the controversy, Stability AI recently partnered with Peter Gabriel to launch a global AI animation challenge called Diffuse Together.
Originally Published 2 years ago — by The Information
Open-source AI models developed by academics using Meta Platforms' machine-learning models are now performing reasonably close to proprietary models from Google and ChatGPT. Most software developers are expected to eventually opt for the free alternatives.