Cargo coding is a style of programming where the developer looks to copy-paste code snippets from somewhere else with the hope that it’ll work. The person at task generally doesn’t know much about the depths of the language and relies on forums like StackOverflow to get their job done.

货物编码是一种编程风格，开发人员希望从其他地方复制粘贴代码片段，并希望它能起作用。 通常，负责任务的人对语言的深度了解不多，并依靠StackOverflow之类的论坛来完成工作。

Now that we’ve got this definition out of the way, let’s focus on the primary subject at hand: OpenAI’s new language model.

现在我们已经摆脱了这个定义，让我们关注手头的主要主题：OpenAI的新语言模型。

Co-founded by the likes of Elon Musk and Peter Theil amongst many others, OpenAI, was once a non-profit organization. It has made huge strides in the past few years by bridging the gap between AI and general human intelligence and is commercial today.

OpenAI由Elon Musk和Peter Theil等人共同创立，曾经是一个非营利组织。 在过去的几年中，它通过弥合AI和一般人类智能之间的鸿沟而取得了长足的进步，并且今天已经商业化。

OpenAI的新GPT-3的炒作是什么？ (What’s the Hype Around OpenAI’s New GPT-3?)

Unless you’re living under a rock, you’d probably have heard about the OpenAI’s latest GPT iteration. It has engulfed the social media by storm.

除非您生活在岩石之下，否则您可能已经听说过OpenAI的最新GPT迭代。 它席卷了社交媒体。

The reason behind GPT-3’s enormous hype right now is due to its mighty model size. Today, it's one of the largest and most powerful machine learning models for natural language processing in the world.

GPT-3目前大肆宣传的原因是其强大的模型尺寸。 如今，它已成为世界上用于自然语言处理的最大，功能最强大的机器学习模型之一。

Trained on over a trillion words and based on a transformer neural network architecture, the Generative Pre-training model is capable of doing almost any natural language task be it writing prompts, QAs for chatbots or even generating language-specific code.

Generative Pre-training模型经过一万亿个单词的训练，基于变压器神经网络体系结构，几乎可以执行任何自然语言任务，无论是编写提示，聊天机器人的QA还是生成特定于语言的代码。

The GPT-3 model uses 175 billion parameters. That’s billion with a B. And its more than 100 times the number of parameters on which GPT-2 was trained.

GPT-3模型使用了1,750亿个参数。 用B表示十亿。它是GPT-2训练参数的100倍以上。

The huge number of parameters and enormous data that are used for training this model has pretty much thrown away any need for fine-tuning — something which BERT and GPT-2 had to fall back on for specific tasks.

用于训练该模型的大量参数和大量数据几乎消除了进行微调的任何需要-BERT和GPT-2必须依靠这些来完成特定任务。

Essentially, you just need to tell the GPT-3 model what needs to be done as the input in English and it’ll compute the output for you based on the next word prediction. For instance, you’d type “If Trump was a programmer” in the input box, and it’ll automatically complete the story for you(until a certain limit of course).

本质上，您只需要告诉GPT-3模型英语需要做些什么作为输入，它就会根据下一个单词的预测为您计算输出。 例如，您在输入框中输入“ If Trump was a一名程序员”，它将自动为您完成该故事(当然要达到一定的限制)。

While currently, OpenAI has been released GPT-3 in private beta only, the early testers have already shared their findings which have went viral on Twitter. One of them showcased how to create a basic web app by simply typing your requirements in plain English and it generates the language code and app for you.

目前，OpenAI仅以私测版发布了GPT-3，早期的测试人员已经分享了他们的发现，这些发现在Twitter上得到了广泛传播。 其中之一展示了如何通过简单地用纯英文输入您的要求来创建基本的Web应用程序，并为您生成语言代码和应用程​​序。

Though there’s a lot to discuss about the intricacies of GPT-3, I’d skip onto what it means for developers and their jobs. If you’re inclined to explore GPT-3 in-depth, I’d recommend this article a read. Or simply Google it.

尽管有很多关于GPT-3复杂性的讨论，但我会跳过对开发人员及其工作的意义。 如果你倾向于探索GPT-3的深入，我建议这个文章一读。 或者只是谷歌它。

程序员，GPT-3不会窃取您的熟练工作 (Programmers, GPT-3 Won’t Steal Your Skilled Jobs)

Since the start of 2020, there has been a pretty common notion that the future of coding is no-code. In fact, there’s a big looming fear that AI would take over humans and their jobs someday.

自2020年初以来，已经有一个非常普遍的观念，即编码的未来就是无代码。 实际上，人们迫切担心人工智能有一天会接管人类及其工作。

That, along with the current rapid emergence of GPT-3 has caused a stir in the developer’s community — with some thinking that the no-code era of programming has finally arrived.

随着当前GPT-3的Swift出现，在开发人员社区引起了轰动-有些人认为编程的无代码时代终于到来了。

While AI could take over most of the programming jobs in a decade from today, the current OpenAI’s GPT-3 model despite its maganamity is still far from general human intelligence.

尽管AI可能会在从今天起的十年内接管大部分编程工作，但当前的OpenAI的GPT-3模型尽管功能强大，但仍远非一般的人类智能。

Programming is a mix of art and science and problem-solving requires some deeper thinking than merely generating code in a visual layout. Moreover, comprehending complex requirements to write production-level code isn’t something AI is ready to do yet in a foolproof way.

编程是艺术与科学的结合，解决问题不仅需要以可视化布局生成代码，还需要更深入的思考。 而且，理解复杂的要求来编写生产级代码并不是AI能够以万无一失的方式完成的。

We already have a lot of no-code visual tools like CreateML for building simple models based on transfer learning. GPT-3 would be nothing but a great addition to that suite to help programmers speed up their workflow.

我们已经有很多无代码的可视化工具，例如CreateML，用于基于迁移学习构建简单的模型。 GPT-3只是该套件的一大补充，可以帮助程序员加快工作流程。

If at all anything, it’ll make the programmer's job more productive as they’d have specific roles devoid of elementary tasks.

如果有的话，这将使程序员的工作更有生产力，因为他们将担任没有基本任务的特定角色。

GPT-3 would never kill jobs skilled developers. Instead its a wake-up call for cargo coders and developers. It’ll urge them to buckle up and upskill to ensure they’re up for solving complex computer programming problems.

GPT-3绝不会杀死熟练的开发人员。 相反，这是对货物编码人员和开发人员的警钟。 它将敦促他们弯腰并提高技能，以确保他们有能力解决复杂的计算机编程问题。

OpenAI’s GPT-3 would help raise the bar of entry-level developers who till now were complacent and happily doing monkey tasks such as copy-pasting code snippets from Google.

OpenAI的GPT-3可以帮助提高入门级开发人员的门槛，这些开发人员迄今仍自满而乐于完成猴子任务，例如Google粘贴粘贴的代码片段。

Perhaps, GPT-3 could decimate the developer’s dependence on Google.

也许，GPT-3可以消除开发人员对Google的依赖。

Also, another positive thing that might emerge from this breakthrough is that it’ll encourage more non-developers to build simple websites and apps themselves whilst gaining programming expertise.

此外，这一突破可能带来的另一件积极的事情是，它将鼓励更多的非开发人员自己构建简单的网站和应用程序，同时获得编程方面的专业知识。

结论 (Conclusion)

Remember, the GPT-3 model, despite its size, is still miles away from human brain intelligence. It’s just a new abstraction layer on programming.

请记住，尽管GPT-3模型的尺寸很大，但与人类大脑的智能仍然相距甚远。 它只是编程上的新抽象层。

Of course, developers who were bound to doing simple no-brainer logic stuff in HTML, CSS, and SwiftUI, might become obsolete as GPT-3 goes commercial in the coming year. But in no way, would it kill the programming or data science jobs.

当然，随着明年GPT-3的商业化，那些不得不在HTML，CSS和SwiftUI上做简单的逻辑工作的开发人员可能会过时。 但这绝不会杀死编程或数据科学工作。

As it is popularly said, the day AI takes over programming, it’ll literally take over all jobs.

众所周知，在AI接管编程的那一天，它将真正接管所有工作。

Elon Musk whose at the helm of OpenAI is actually looking to mitigate AI’s threat to humanity. His vision is to blend Artificial Intelligence with us in such a way that we have the ultimate control over it and not the other way round. GPT-3 might not be close to general artificial intelligence yet, but its perhaps the first stepping stone where AI meets the common man.

埃隆·马斯克(Elon Musk)，他是OpenAI的负责人，实际上是在寻求减轻AI对人类的威胁。 他的愿景是将人工智能与我们融为一体，从而使我们对其拥有最终控制权，而不是相反。 GPT-3可能尚未接近通用人工智能，但它可能是AI与普通人相遇的第一个垫脚石。