AI for technical writing
More technical writers use AI to help them in their daily work. It's time to embrace AI as a technical writing tool.
News about the potentially transformative capabilities of Artificial Intelligence (AI) are everywhere. Opinions on the topic range from "AI will lead us into a technological paradise" to "AI will put us all out of work." As a technical writer, should you be worried?
Let's explore these AI capabilities based on how they complement your skills as a technical writer. But first, let's make sure we have a common understanding of what AI really is.
What is AI
Artificial intelligence or machine learning (ML) refers to the simulation of human intelligence in machines that are programmed to perform tasks that typically require human intelligence. These tasks include learning, reasoning, problem-solving, perception, understanding natural language, and decision-making. The primary goal of AI is to create machines or systems that can mimic or replicate cognitive abilities that are typically associated with human intelligence.
AI can be classified into two main categories:
Narrow AI (Weak AI) - This type of AI is designed and trained for a specific task or set of tasks. It can excel at performing those particular tasks, but it lacks general cognitive abilities and cannot transfer its knowledge to other domains. Examples of narrow AI include virtual assistants like Siri or Alexa, recommendation systems on e-commerce websites, and image recognition algorithms.
General AI (Strong AI) - This refers to AI that possesses the ability to understand, learn, and apply knowledge across a wide range of tasks similar to human intelligence. General AI would have cognitive abilities that allow it to reason, solve problems, and learn from experiences in various domains. However, achieving true general AI remains a challenging and theoretical goal, as it requires replicating the complexity and adaptability of the human mind.
Artificial intelligence can be achieved using different techniques, such as machine learning, deep learning, natural language processing (NLP), and computer vision. These techniques involve training algorithms on vast amounts of data to identify patterns, make predictions, and improve their performance over time.
AI has the potential to revolutionize various industries, including healthcare, finance, transportation, manufacturing, and more. It also raises important ethical and societal considerations, such as concerns about job displacement, privacy, bias in algorithms, and the responsible use of AI technology. As AI continues to evolve, researchers and policymakers strive to strike a balance between innovation and addressing these important challenges.
Prompt engineering and chaining
We're already seeing the emergence of a specialized skill set around writing effective prompts for AI to efficiently retrieve a response in the necessary format, length, writing style, and so on. In some cases, information may need to be provided with the prompt to create a useful response. Often, experimentation and iteration will be needed to achieve the ideal response.
There's also a concept of chaining: automating the generation of prompts, often by using software to use the output of one prompt to create new prompts, typically to achieve more complex results. For example, a colleague of mine uses ChatGPT to create image generation prompts that he then feeds into the Midjourney engine, to get better results than if he were to write those prompts himself. The LangChain framework has become a popular tool to address more complex use cases, by automating a sequence of prompts, often spanning multiple Large Language Models (LLMs), such as ChatGPT, Hugging Face Hub, and more.
It's time to embrace AI as a productivity tool
AI tools continue to evolve quickly. The recent release of ChatGPT 4 has been hailed by many as a major step forward in terms of the quality of content it could produce. Still, the limitations on the recency of information it has available continue. For now.
Meanwhile, efforts are underway to make training new LLMs easier than ever before. Infrastructure solutions like Amazon Sagemaker aim to reduce the technical expertise needed to create and train ML models, so we start to see more focused, domain-specific models useful in more technical areas. An early example is Github's Copilot, which has been trained to help programmers write code.
Today, AI can be an excellent way to amp your productivity, and allow you to do a better job at the work you love doing. If you can embrace the tools and start building the muscle memory of leveraging these tools in your daily work, a future where AI is embraced more broadly will only enhance your technical writing prospects.
Martin Anderson-Clutz is a Senior Solutions Engineer at Acquia.