Facilitating Learning Part 2: The Rise of AI 《促进学习 》第二部分：AI的崛起

AI is more familiar than you might think. To start, Alan Tuning postulated AI in the early 40’s, and the “Turning Test” is still our best benchmark for evaluating AI frameworks. However, at this time, AI was more science fiction than science fact. An academic curiosity at most; That said, functional AI models such as Chat GPT are more familiar than you might think.

人工智能比你想象的要熟悉得多。首先，艾伦-特宁（Alan Turning）在 40 年代初就提出了人工智能的概念，”特宁测试 “至今仍是我们评估人工智能框架的最佳基准。然而在当时，人工智能与其说是科学事实，不如说是科幻小说。尽管如此，像 Chat GPT 这样的人工智能功能模型比你想象的要熟悉得多。

Back in the day, I had the famed Pentium 486 Computer, which was running MS-DOS. I eventually upgraded the operating system to Windows 98, and at this time, I also installed an early AI SDK (Software Developer Kit). While these early software prototypes were nowhere near as advanced as they are now, I could still generate a query and the computer could respond to my request and complete simple tasks on my behalf. However, the AI was very limited as it was bound by the local database installed locally on the computer and could not communicate with the outside world. In contrast to these early prototypes, newer AI algorithms now have access to global databases — and that has made them exponentially more powerful and arguably more valuable and helpful than they have ever been in the past thirty years. 

当年，我有一台著名的奔腾 486 电脑，运行的是 MS-DOS。最终，我将操作系统升级到了 Windows 98，同时还安装了一个早期的人工智能 SDK（软件开发工具包）。虽然这些早期的软件原型远没有现在这么先进，但我仍然可以生成一个查询系统，计算机可以响应我的请求并代表我完成简单的任务。不过，人工智能的功能非常有限，因为它受到安装在计算机的本地数据库的限制，无法与外界交流。与这些早期的原型相比，现在较新的人工智能算法可以访问全球数据库–这使它们的功能成倍增加，可以说比过去三十年中的任何时候都更有价值和帮助。

However, to someone like me, I look at these advancements with little to no reaction. I am somewhat jaded by my experiences watching them evolve over the decades. I just look at what we have today and say… “Yep, that is a logical jump in the technology,” but I am not mystified like most people viewing the technology as if they are seeing magic for the first time. After all, Arthur C. Clark once said, “Any sufficiently advanced technology is indistinguishable from magic!” Furthermore, that is what AI is for most people! Indistinguishable from magic!

然而，对于像我这样的人来说，我对这些进步知之甚少。几十年来，我目睹了这些技术的发展，因此有些厌倦。我只是看着我们今天所拥有的一切，然后说…… “是的，这是技术上的一个合乎逻辑的飞跃，”但我并不像大多数人那样对技术感到神秘，就像他们第一次看到魔法一样。毕竟，阿瑟-C-克拉克曾经说过：”任何足够先进的技术都与魔法无异！”此外，对大多数人来说，人工智能就是如此！与魔法无异！

Like any innovation, you have cutting-edge research being done, and over time, aspects of these advanced research projects slowly trickle down into our mainstream technologies. So, what did the subtle changes in our everyday technology that led us to where are today look like? After all, Chat GPT can query a simple prompt against a vast global database of networked computers representing the sum total of human knowledge and return a seemingly original response to a user’s prompt. How did we get to this point? Contrary to most people’s belief, this was not a sudden evolutionary leap in technology. Instead, it has been a slow and evolving process.

就像任何创新一样，前沿研究一直在进行，随着时间的推移，这些先进研究项目的各个方面会慢慢渗透到我们的主流技术中。那么，我们的日常技术究竟发生了哪些微妙的变化，导致我们取得了今天的成就？毕竟，Chat GPT 可以根据代表人类知识总和的庞大全球联网计算机数据库查询一个简单的提示，并对用户的提示返回一个看似原创的响应。我们是如何做到这一点的？与大多数人的想法相反，这并不是技术发展的突然飞跃。相反，这是一个缓慢而不断发展的过程。

The first mainstream adoption of this technology can be seen in the release of the early spell checkers on the personal computer in 1980. This innovation, the great-grandparent of modern-day AI, checked the spelling of words against a known database. It was a simple algorithm that queried the spelling of words against a locally installed database, and these simple algorithms helped users to create content in the digital age. This technology then evolved into grammar-checking technologies. Much like their predecessor, grammar checks used more complex mathematical algorithms to analyze the structure of a statement. They identified misplaced or inaccurate word choices based on the part of speech, tense, and grammatical context of each word in relationship to the other words that were used in the sentence. Again, this advancement in technology-assisted users to develop better content in the digital era, and this was thanks to the continued advancements that were being made in computer algorithm design.

1980 年，个人电脑发布了早期的拼写检查程序，这是这项技术首次被主流采用。这项创新是现代人工智能的曾祖，它根据已知数据库检查单词拼写。这是一种根据本地安装的数据库查询单词拼写的简单算法，这些简单的算法帮助用户在数字时代创建内容。这项技术随后发展成为语法检查技术。与前者一样，语法检查使用更复杂的数学算法来分析语句的结构。它们根据每个单词的语篇、时态和语法上下文与句子中其他单词的关系来识别错位或不准确的单词选择。技术的进步再次帮助用户在数字时代开发出更好的内容，而这要归功于计算机算法设计的不断进步。

This 1980s innovation still seems a long way from what we have today, but like everything else, the technology continued to evolve slowly. The next significant advancement was the “Mail Merge.” Using mail merge technologies, customized content could be created for an infinite number of situations based on standard information fields such as <<name>>, <<date>>, <<address>>, and so forth. This technology was commonly used by businesses to generate letters or reports for clients, and it was also used heavily by teachers when writing report cards (e.g., PowerSchool comment banks). For instance, the teacher only needed to select the comment they wanted to use, and the computer automatically adjusted the comment by adding the student’s name and adjusting gender-based pronouns throughout the comment (i.e., he/she, his/her). Therefore, these new advancements in software and algorithm design saved teachers a massive amount of time and reduced errors that could result in generating descriptive comments about their students. Again, the technology made creating content in the digital era even more accessible than it had been with the previous iterations of the technology.

这一 20 世纪 80 年代的创新与我们今天所拥有的技术相比，似乎仍有很大差距，但就像其他任何事物一样，技术仍在缓慢发展。下一个重大进步是 “邮件合并”。利用邮件合并技术，可以根据<<姓名>>、<<日期>>、<<地址>>等标准信息字段创建无限量的自定义内容。企业通常使用这种技术为客户生成信件或报告，教师在撰写成绩单（如 PowerSchool 评语库）时也大量使用这种技术。例如，教师只需选择要使用的评语，计算机就会自动调整评语，在评语中加入学生的姓名并调整基于性别的代词（如他/她、他/她的）。因此，这些软件和算法设计方面的新进展为教师节省了大量时间，减少了在生成有关学生的描述性评论时可能出现的错误。同样，与之前的技术迭代相比，该技术使数字时代的内容创建更加便捷。

Taking the technology even further, we started seeing predictive writing prompts introduced into quick messaging services. This new technological advancement would make automated predictions about what word you would likely use next to help you type messages quickly and easily. All this was made possible thanks to advancing computer algorithms that were able to quickly break down what you were writing in real-time into a mathematical sequence that represented a grammatical structure within the target language. The software would then use standard grammar models to predict the next word you would need based on your written content.

随着技术的进一步发展，我们开始看到快速信息服务中引入了预测性写作提示。这项新的技术进步会自动预测你下一步可能使用的单词，帮助你快速、轻松地输入信息。这一切都要归功于不断进步的计算机算法，它们能够将您实时书写的内容快速分解成代表目标语言语法结构的数学序列。然后，软件会使用标准语法模型，根据您所写的内容预测您需要的下一个单词。

Predictive writing tools were then taken even further by companies like “Grammarly” and “Speechly.” These, the closest in kin to AIs like Chat GPT, were able to assist users in writing content quickly while ensuring the content had the right tone, consistency, and accuracy. However, unlike predictive writing tools, which could assist in individual word choices, these more advanced algorithms were able to rewrite entire sentences while keeping the intended meaning the same. That said, these emerging technologies still needed to be improved in the sense that they needed an operator to guide and direct them in the content creation process. They could not create content on their own, and that is how AI’s programs, such as Chat GPT, changed everything as it can create content entirely on its own based on an initial user query — and that is a real game changer.

随后，”Grammarly “和 “Speechly “等公司进一步推出了预测性写作工具。这些工具与 Chat GPT 等人工智能最为接近，能够帮助用户快速撰写内容，同时确保内容具有正确的语气、一致性和准确性。不过，预测性写作工具可以帮助用户选择个别词语，而这些更先进的算法则不同，它们可以重写整个句子，同时保持原意不变。尽管如此，这些新兴技术仍需改进，因为它们在内容创建过程中需要操作员的指导和引导。它们无法独立创建内容，而人工智能程序（如 Chat GPT）就是这样改变了一切，因为它可以根据用户的初始查询完全独立地创建内容–这才是真正改变了游戏规则。

Therefore, in summary, as our computer algorithms become more and more sophisticated, so are the tasks in which they can complete. So much so that many people now fear for their jobs, and it has upset a delicate balance in the educational sector. Moreover, according to a recent Forbes report, AI will replace close to 85 million jobs by the year 2025, and that is not that far away, but what type of jobs are most at risk, though? Regrettably, most of them will be the coveted high-paying jobs that many university graduates fight over. Jobs like my first data analytics job which will be automated entirely going forward. 

因此，综上所述，随着我们的计算机算法越来越复杂，它们所能完成的任务也越来越多。以至于现在很多人都担心自己的工作，这也打破了教育领域的微妙平衡。此外，根据《福布斯》最近的一份报告，到 2025 年，人工智能将取代近 8500 万个工作岗位，而这并不遥远。令人遗憾的是，其中大部分将是许多大学毕业生梦寐以求的高薪工作。像我的第一份数据分析工作，未来将完全自动化。

Is this the end of education as we know it?

这就是我们所知的教育的终结吗？ 

Are we all going to be replaced by machines? 

我们都将被机器取代吗？

Moreover, what type of Jobs will be left in the post-AI era? 

此外，后人工智能时代会留下什么样的乔布斯？

Well, in my opinion, it will be the jobs that the school system has not been doing a great job preparing people for, which will survive the AI revolution. Jobs in the fields of STEM where you must use your hands in addition to your brain: Jobs that use the Engineering & Design Process and the Scientific Method. Things that a computer cannot do! These are the jobs that involve CRITICAL THINKING in combination with applied (hands-on) skill sets. However, jobs that are purely analytical in nature will be entirely replaced by computers in the coming years.

在我看来，这是因为那些学校系统没有很好地培养人才的工作，它们很难在人工智能革命中幸存下来。STEM领域的工作，除了动脑，还必须动手：使用工程设计流程和科学方法的工作。电脑无法完成的工作！这些工作需要结合应用（动手）技能进行批判性思考。然而，在未来几年内，纯粹分析性的工作将完全被计算机所取代。

Now you may ask, why is this? And the answer is quite simple. Many coveted white-collar jobs are focused solely on data analytics. Something that a computer is uniquely qualified for. For years, we have trained and conditioned humans to be more like computers; all the while, we have been creating algorithms to allow computers to function more like people. Now, we are on an uneven playing field, and there is no way humans can ever compete with the computational power of a modern-day computer. That said, there is something that computers still cannot do, and that is to interact with the physical world around them. Therefore, jobs that require people to use their hands… careers in the fields of STEM, will be the future of education. However, that is nothing new. We have known this since the National Science Foundation introduced STEM back in 2001. The only difference is that we can no longer ignore the warning signs. Change is at our doorstep whether we like it or not, and we will be left with two choices — change with the times or be left behind!

现在你可能会问，这是为什么呢？答案很简单。许多令人羡慕的白领工作只专注于数据分析。而这正是电脑所能胜任的。多年来，我们一直在训练人类，让他们更像电脑；同时，我们也一直在创造算法，让电脑的功能更像人。现在，我们处在一个不公平的竞争环境中，人类根本无法与现代计算机的计算能力相抗衡。尽管如此，计算机仍然无法做到与周围的物理世界进行交互。因此，需要人们动手的工作……STEM 教育领域的职业，将是教育的未来。然而，这并不是什么新鲜事。早在 2001 年美国国家科学基金会提出 STEM教育 时，我们就已经知道了这一点。唯一不同的是，我们再也不能忽视这些警示信号了。无论我们愿不愿意，变革都已经来到了我们的家门口，我们将面临两个选择–与时俱进或被时代抛在后面！