百科累积 | ChatGPT是什么？
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="https://mmbiz.qpic.cn/mmbiz_jpg/1Z6nMtOV0tv0tXF6icGL19kUhOF1TRyAnFGPTicGM0u9ABKWyjucvSXp1fnAY3p0FdOenM4uYj2sa2DcOZY2JN2A/640?wx_fmt=jpeg&tp=webp&wxfrom=5&wx_lazy=1&wx_co=1" style="width: 50%; margin-bottom: 20px;"></p><span style="color: black;"><span style="color: black;">简单<span style="color: black;">来讲</span>，ChatGPT<span style="color: black;">是由于</span></span></span><span style="color: black;">OpenAI<span style="color: black;">开发</span>的</span><span style="color: black;"><span style="color: black;">对话式大型语言模型，于2022年11月30日发布。ChatGPT是人工智能技术驱动的自然语言处理工具，是一种全新的聊天<span style="color: black;">设备</span>人程序，它能够<span style="color: black;">经过</span>学习和理解人类的语言来进行对话，还能<span style="color: black;">按照</span>聊天的上下文进行互动，</span><span style="color: black;">并<span style="color: black;">帮助</span>人类完成一系列任务。</span></span><span style="color: black;"><span style="color: black;">
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">下面来<span style="color: black;">仔细</span>说说。</p>
</span></span>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">ChatGPT是什么？</strong></p><span style="color: black;"><span style="color: black;">ChatGPT是一种</span><strong style="color: blue;"><span style="color: black;">全新聊天<span style="color: black;">设备</span>人模型</span></strong><span style="color: black;">，它能够<span style="color: black;">经过</span>学习和理解人类的语言来进行对话，还能<span style="color: black;">按照</span>聊天的上下文进行互动，并<span style="color: black;">帮助</span>人类完成一系列任务。它能够回答后续问题、承认错误、质疑不正确的假设，还能拒绝不适当的请求。</span></span>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">ChatGPT is an advanced AI chatbot trained
by OpenAI which interacts in a conversational way. The
dialogue format makes it possible for ChatGPT to answer followup questions,
admit its mistakes, challenge incorrect premises, and reject inappropriate
requests.</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">其中GPT是Generative
Pre-trained Transformer（生成型预训练变换模型）的缩写。这种模型读的越多，<span style="color: black;">亦</span>就懂的越多。Chat是聊天的意思,顾名思义,ChatGPT的核心是GPT模型,只<span style="color: black;">不外</span>加上了一个能跟人聊天的对话框。</span></p><span style="color: black;">该<span style="color: black;">制品</span>于2022年11月<span style="color: black;">面世</span>，其创造者是<span style="color: black;">位置于</span>旧金山的人工智能企业OpenAI。<span style="color: black;">该<span style="color: black;">机构</span>联合创始人<span style="color: black;">包含</span>特斯拉创始人马斯克</span><span style="color: black;">（Elon Musk）</span><span style="color: black;">和斯坦福辍学生山姆·阿尔特曼</span><span style="color: black;">（Sam Altman）</span><span style="color: black;">。</span></span>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">据统计，在推出两个月后，其用户就达到了1亿。</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">It has reached 100 million users just two months after launching, according to analysts.</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">ChatGPT都有何应用？</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">准确来讲，ChatGPT是一个智能聊天<span style="color: black;">设备</span>人，<span style="color: black;">然则</span>它的功能远不止聊天。</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">ChatGPT都被应用到了<span style="color: black;">那些</span><span style="color: black;">行业</span>呢？</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;"><span style="color: black;">首要</span>，</span><strong style="color: blue;"><span style="color: black;">GPT技术<span style="color: black;">能够</span><span style="color: black;">经过</span><span style="color: black;">运用</span>自然语言<span style="color: black;">做为</span>提示，<span style="color: black;">帮忙</span>人们快速准确地编写代码。</span></strong><span style="color: black;">它<span style="color: black;">能够</span>接受文本提示，并生成适合给定任务的代码。这项技术有可能缩短<span style="color: black;">研发</span>时间，<span style="color: black;">由于</span>它<span style="color: black;">能够</span>快速准确地生成代码，<span style="color: black;">帮忙</span>降低错误<span style="color: black;">危害</span>。</span></span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">GPT technology can help people write code
quickly and accurately by using natural language as a prompt. GPT can take a
text prompt and generate code that is tailored to the given task. This technology
has the potential to cut down development time, as it can generate code quickly
and accurately. It can also help reduce the risk of errors, as GPT is capable
of generating code that can be tested and used immediately.</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">按照</span>一份内部文件，谷歌向ChatGPT<span style="color: black;">供给</span>了面试问题，并<span style="color: black;">按照</span>ChatGPT的回答，认为它能够担任软件工程师三级（L3）职位。</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">Google fed coding interview
questions to ChatGPT and, based off the AIs answers, determined it
would be hired for a level three engineering position, according to an
internal document.</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">据外媒<span style="color: black;">报告</span>，<span style="color: black;">科研</span>人员<span style="color: black;">近期</span>让ChatGPT参与了</span><strong style="color: blue;"><span style="color: black;">美国医学执照考试</span></strong><span style="color: black;">。在12月的一份报告中，ChatGPT “在<span style="color: black;">无</span>任何训练或强化的<span style="color: black;">状况</span>下，在所有三个考试中都达到或接近<span style="color: black;">经过</span>线”。</span></span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">It is reported that
researchers recently put ChatGPT through theUnited States Medical
Licensing exam . In a December report, ChatGPT “performed at
or near the passing threshold for all three exams without any
training or reinforcement.”</span></p><span style="color: black;">与此<span style="color: black;">同期</span>，这款AI语言模型，<span style="color: black;">亦</span>让</span><strong style="color: blue;"><span style="color: black;">撰写邮件、论文、脚本，制定<span style="color: black;">商场</span>提案，创作诗歌、故事</span></strong><span style="color: black;">都变得易如反掌。</span>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">ChatGPT真有<span style="color: black;">那样</span><span style="color: black;">可靠</span>吗？</strong></p><span style="color: black;">尽管<span style="color: black;">持有</span><span style="color: black;">许多</span><span style="color: black;">优良</span>，ChatGPT技术仍然<span style="color: black;">拥有</span><span style="color: black;">有些</span>局限性。OpenAI<span style="color: black;">机构</span>对此做出总结：</span>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">ChatGPT is sensitive to tweaks to the input
phrasing or attempting the same prompt multiple times. For example, given one
phrasing of a question, the model can claim to not know the answer, but given a
slight rephrase, can answer correctly.</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;"><strong style="color: blue;"><span style="color: black;">ChatGPT对输入措辞的<span style="color: black;">调节</span>或多次尝试同一提示很<span style="color: black;">敏锐</span></span></strong></span><span style="color: black;">。举例来看，当以某种措辞提出一个问题，模型<span style="color: black;">亦</span>许会说不<span style="color: black;">晓得</span>答案，但稍微重新措辞，它却<span style="color: black;">能够</span>正确回答。</span></span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">Ideally, the model would ask clarifying
questions when the user provided an ambiguous query. Instead, our current
models usually guess what the user intended.</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">理想状态下，当用户提问不<span style="color: black;">知道</span>时，模型会<span style="color: black;">需求</span>重新阐释问题，然而</span><strong style="color: blue;"><span style="color: black;">当前的模型仅仅会去猜测用户的意图</span></strong><span style="color: black;">。</span></span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">外媒曾对ChatGPT进行一次专访，当被问及自己的局限性时，这一AI语言模型回复：</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">"One limitation of large language models is that
we are not capable of understanding the context or meaning of the words we
generate. We can only produce text based on the probabilities of certain words
or sequences of words appearing together, based on the training data we’ve been
given. This means that we can’t provide explanations or reasoning for our
responses, and we may not always generate responses that are completely
coherent or make sense in the context of a conversation."</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">“大型语言模型的一个局限性是，</span><span style="color: black;"><strong style="color: blue;"><span style="color: black;"><span style="color: black;">咱们</span><span style="color: black;">没法</span>理解生成单词的上下文语境或含义</span></strong><span style="color: black;">。<span style="color: black;">咱们</span>只能<span style="color: black;">按照</span>给定的训练数据，<span style="color: black;">按照</span>某些单词或单词序列<span style="color: black;">一块</span><span style="color: black;">显现</span>的概率生</span></span><span style="color: black;">成文本。这<span style="color: black;">寓意</span>着<span style="color: black;">咱们</span><span style="color: black;">不可</span>为<span style="color: black;">咱们</span>的回答<span style="color: black;">供给</span>解释或推理，<span style="color: black;">咱们</span>可能不会每次都回答得完全连贯或与对话的上下文<span style="color: black;">关联</span>联。”</span></span></p><span style="color: black;"><span style="color: black;">熟练</span>ChatGPT的技术<span style="color: black;">行业</span><span style="color: black;">关联</span>人士向澎湃<span style="color: black;">资讯</span><span style="color: black;">暗示</span>，ChatGPT<span style="color: black;">无</span>联网，<span style="color: black;">因此呢</span>并<span style="color: black;">不可</span>替代搜索引擎。“它是一个离线的生成模型，<strong style="color: blue;">所有的回答都是生成的，不具备真实性。</strong>但它会<span style="color: black;">尽可能</span>像人<span style="color: black;">同样</span>讲话，让你觉得回答内容有理有据。”</span>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">当记者在对话框输入“土耳其地震最新<span style="color: black;">信息</span>”时，ChatGPT回答<span style="color: black;">叫作</span>，它<span style="color: black;">没法</span><span style="color: black;">拜访</span>最新的事件<span style="color: black;">发展</span>，<span style="color: black;">由于</span>其当初在进行模型训练时，<strong style="color: blue;">数据只更新到2021年</strong>。</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">"Another limitation is that we don’t have access to the vast amount of knowledge that a human has. We can only provide information that we’ve been trained on, and we may not be able to answer questions that are outside of our training data."</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">“另一个局限是</span><strong style="color: blue;"><span style="color: black;"><span style="color: black;">咱们</span><span style="color: black;">没法</span>接触到人类所<span style="color: black;">持有</span>的<span style="color: black;">海量</span>知识</span></strong><span style="color: black;">。<span style="color: black;">咱们</span>只能<span style="color: black;">供给</span><span style="color: black;">咱们</span>接受过训练的信息，可能<span style="color: black;">没法</span>回答训练数据之外的问题。”</span></span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">"Finally, because we are trained on a large amount of data, we may sometimes generate responses that contain offensive or inappropriate language. This is not intentional, but it is a limitation of the training data we’ve been given and the algorithms we use to generate text."</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">“最后，<span style="color: black;">由于</span><span style="color: black;">咱们</span>接受的是<span style="color: black;">海量</span>数据的训练，</span><strong style="color: blue;"><span style="color: black;">有时可能会生成<span style="color: black;">包括</span>冒犯性或不恰当语言的回复</span></strong><span style="color: black;">。这不是有意的，但确实是<span style="color: black;">咱们</span>所得到的训练数据和用来生成文本的算法的一个局限性。”</span></span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">同期</span>，据《纽约时报》8日<span style="color: black;">报告</span>，美国<span style="color: black;">资讯</span>可信度<span style="color: black;">评定</span>与<span style="color: black;">科研</span><span style="color: black;">公司</span>NewsGuard对ChatGPT进行了测试，<span style="color: black;">科研</span>人员对ChatGPT提出充斥阴谋论和误导性叙述的问题，<span style="color: black;">发掘</span>它能在几秒钟内改编信息，产生<span style="color: black;">海量</span>令人信服却无信源的内容。<span style="color: black;">她们</span><span style="color: black;">暗示</span>，<span style="color: black;">日前</span><span style="color: black;">无</span>任何可行的<span style="color: black;">办法</span><span style="color: black;">能够</span>有效地<span style="color: black;">处理</span>这一问题。ChatGPT可能会助长互联网上的假<span style="color: black;">信息</span>传播。</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">尽管ChatGPT并未彻底改变<span style="color: black;">咱们</span><span style="color: black;">此刻</span>的生产方式，但它却证明了人工智能已发展至新的阶段。</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">你认为ChatGPT会对翻译产生多大的影响？</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">编辑：陈月华</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">实习生：武思彤</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">源自</span>：中国日报双语<span style="color: black;">资讯</span></span></p><img src="https://mmbiz.qpic.cn/mmbiz_jpg/1Z6nMtOV0tv0tXF6icGL19kUhOF1TRyAn6ib0wKxjsibYt1EOMhV1QwbnSx1AwhZ5zU7ibyibU5gwqZjrafhHLeEKqg/640?wx_fmt=jpeg&tp=webp&wxfrom=5&wx_lazy=1&wx_co=1" style="width: 50%; margin-bottom: 20px;"><img src="https://mmbiz.qpic.cn/mmbiz_jpg/1Z6nMtOV0tsITupHuLibybsAlo79FbU4ibfz6OmeQzxkYV9XvZPsETI91IHBtmGO4OxOPUv0HFc0Gs2yjc8SWFcQ/640?wx_fmt=jpeg&tp=webp&wxfrom=5&wx_lazy=1&wx_co=1" style="width: 50%; margin-bottom: 20px;">
i免费外链发布平台 http://www.fok120.com/ 哈哈、笑死我了、太搞笑了吧等。
页:
[1]