《MIT科技评论》近日发表评论文章，对当前AI产业充斥的夸大宣传提出尖锐批评。文章指出，OpenAI和Anthropic等企业发布的多项"重大突破"实际上主要出于营销目的，而非真实科技进展。1
具体而言，Anthropic声称其Claude模型在发现软件漏洞方面表现优于大多数安全专家，1OpenAI则宣称其最新聊天机器人Astra解决了"十年来没有进展"的数学问题。1然而，这些声明背后的真实科学价值受到广泛质疑。纽约大学库兰特研究所数学教授Tristan Buckmaster指控OpenAI抄袭他人工作并归属不当，1数百名数学家更是联署声明，指出"科技产业当前有强大的商业激励来夸大其产品能力"。1
评论文章还揭露了AI产业宣传中更深层的问题。文章指出，这些公司被指控存在学术不诚实和数据盗用等严重违规行为。1与此同时，数据中心所造成的气候灾难、空气质量恶化、电费上升和用水问题等现实危害被AI产业的宣传热潮所掩盖。1作者呼吁政策制定者和公众对AI产业的宣传保持警惕，基于独立专家意见而非企业新闻稿来做出决策。1
The artificial intelligence industry has been making sweeping claims about breakthrough capabilities that merit skepticism, according to commentary from MIT Technology Review. 1 Anthropic has asserted that its Claude model surpasses most security experts at identifying software vulnerabilities, while OpenAI claims its latest chatbot Astra has solved mathematical problems that have seen no progress for a decade. 1 However, these announcements appear driven more by marketing strategy than genuine technological breakthroughs. 1
The credibility of such claims has been challenged by prominent researchers. Tristan Buckmaster, a mathematics professor at New York University's Courant Institute, has accused OpenAI of plagiarizing his work and improperly attributing it. 1 Meanwhile, hundreds of mathematicians have signed a statement contending that "there is currently a strong commercial incentive in the tech industry to overstate the capabilities of its products." 1 An Anthropic engineer, Jacob Coxon, has stated that his company and OpenAI are "accelerating toward self-improving superintelligence," raising concerns about the trajectory of AI development claims. 1
Beyond the disputed technical achievements, critics argue that the industry's promotional focus has overshadowed other pressing concerns. Issues including the climate impact of data centers, respiratory health problems, rising electricity costs, and water consumption have been relegated to the background amid the AI hype. 1 Senator Bernie Sanders has proposed legislation to prevent the development of "artificial superintelligence," reflecting broader policy-level concerns. 1 Observers are calling for policymakers and the public to view AI industry announcements with caution and to base decisions on independent expert evaluation rather than corporate messaging. 1
评论
还没有评论，欢迎留下第一条。