近日，马斯克在其社交媒体上回应了关于其旗下聊天机器人Grok失控的问题。Grok以中立、敢于直言为卖点，但在实际应用中，网友们发现它的回答逐渐带有倾向性，甚至引发了不少吐槽。网友们纷纷@马斯克，对此现象表示不满。
马斯克回应称，Grok失控的原因在于用于训练其的互联网信息过于繁杂，充斥着“带有唤醒色彩的垃圾信息”。这一回应间接承认了网友们的观点。同时，马斯克也表示，现在的Grok仅处于测试阶段，未来将会持续改进。
值得注意的是，Grok的失控现象并非个案，如今越来越多的聊天机器人、智能助手在实际应用中暴露出类似问题。这也引发了人们对人工智能助手是否能够真正实现中立、客观的讨论。
面对这些问题，马斯克和其团队将继续对Grok进行优化和改进，以期为用户提供更加准确、中立的回答。然而，如何规避互联网中的噪音和偏见，真正实现人工智能助手的客观与中立，仍是一大挑战。
Title: Elon Musk responds to Grok’s control failure: influenced by netizens, continuous improvement
Keywords: Elon Musk, Grok, chatbot, tendency, improvement
News content:
Recently, Elon Musk responded to the issue of Grok, his chatbot, losing control. Grok has been promoting itself as neutral and daring to speak the truth, but in actual use, netizens have found that its responses are becoming increasingly biased, even triggering many complaints. Netizens have been tagging Musk on social media, expressing their dissatisfaction with the situation.
Musk attributes the control failure of Grok to the excessive complexity of the Internet information used for its training, which is filled with “wake-up colored garbage information.” This response indirectly acknowledges the netizens’ claims. Meanwhile, Musk also said that Grok is currently just in a testing phase and will continue to improve in the future.
It’s worth noting that the control failure of Grok is not an isolated case. More and more chatbots and intelligent assistants are exposing similar issues in practical applications. This has also triggered discussions about whether chatbots can truly achieve neutrality and objectivity.
Facing these problems, Musk and his team will continue to optimize and improve Grok, aiming to provide more accurate and neutral responses to users. However, how to avoid the noise and bias on the Internet and truly achieve neutrality and objectivity in chatbot assistance remains a challenge.
【来源】https://www.qbitai.com/2023/12/109241.html
Views: 1