I think there are broadly three areas in which Generative AI is having an impact on creativity and/or productivity at work. I asked ChatGPT to elaborate on the three and the result I got is below. A great result if I may say and far better than I could have done in the time. So a perfect use case for use of AI to improve my productivity. It’s in case of the area covered in the last column, writing, that I’ll focus on in this post, in relation to the post title. Mostly because I have most experience of its use in this area.

⚠️ NOTE ⚠️ I’m writing on this subject as a way to explore the topic for an upcoming trend report I am working on (Future of HumAIn Work) as well as because I work in this area with customers at the moment (disclosure) so it’s a great way for me to sensemake.

Images

1.Image Classification: This is the task of categorizing an image into one of several predefined classes. For instance, determining if a given picture is of a cat, dog, or bird.

2.Object Detection: AI can identify and locate multiple objects within a single image. This is widely used in applications such as surveillance, autonomous vehicles, and retail analytics.

3.Image Segmentation: This is the task of dividing an image into multiple segments, each of which corresponds to an object or a part of an object. It’s useful in medical imaging, for example, to segment different tissues, tumors, etc.

4.Face Recognition: AI can identify individuals from their facial features. This has applications in security, social media, and many other areas.

5.Style Transfer: AI can apply the artistic style of one image to transform another image. This technique became popular for creating unique and artistic versions of everyday photographs.

6.Image Generation: Generative models, like GANs (Generative Adversarial Networks), can create entirely new images, which can be used for art, video game design, or even fashion.

7.Image Enhancement: AI can improve image quality by increasing resolution (super-resolution), reducing noise, or enhancing details that are not visible to the human eye.

8.Image Captioning: AI models can generate descriptive captions for images, which can be useful for the visually impaired or for content indexing.

9.Medical Imaging: AI aids in detecting diseases by analyzing X-rays, MRIs, and other types of medical images. It can spot patterns that might be overlooked by human professionals.

10.Augmented Reality (AR): AI can recognize and track objects in real time to overlay digital information on the real world.

11.Image Compression: Using AI, we can achieve better image compression rates without compromising much on quality.

12.Visual Search: Instead of text queries, users can search databases or the internet using images as queries.

13.Anomaly Detection: In industrial scenarios, AI can spot defects or anomalies in products through image analysis.

14.Gesture Recognition: AI can recognize and interpret human gestures from images or video streams.

15.3D Reconstruction: From a series of 2D images, AI can reconstruct 3D models. This has applications in areas like archaeology, real estate, and gaming.

Coding

1.Code Completion and Suggestion: Tools like GitHub Copilot, which is powered by OpenAI’s Codex, can automatically suggest whole lines or blocks of code as a developer types, making the coding process faster and more efficient.

2.Bug Detection: AI can analyze code to detect potential bugs, vulnerabilities, or code smells, suggesting fixes even before the code is run or compiled.

3.Code Optimization: AI can suggest performance improvements by analyzing the code structure and understanding bottlenecks.

4.Automatic Documentation: AI can generate comments and documentation for code based on its understanding of the code’s functionality.

5.Code Translation: Convert code from one programming language to another. While not perfect, this can be a starting point for developers porting software.

6.Natural Language to Code: With AI, it’s possible to convert natural language commands into functional code. This can be particularly useful for people without a deep coding background.

7.Test Generation: AI can analyze software and generate test cases that ensure maximum code coverage, helping in rigorous software testing.

8.Code Search: AI can help developers find relevant code snippets or projects that are similar to what they are working on or provide solutions to a particular problem.

9.UI Design and Prototyping: AI can transform a hand-drawn design or a verbal description into a working prototype or even a functional application.

10.Learning and Training: AI-driven platforms adapt and create personalized learning paths for budding coders, focusing on their strengths and weaknesses.

11.Infrastructure Management: In the domain of DevOps, AI can optimize infrastructure resources, predict failures, and automate repetitive tasks.

12.Security: AI can monitor application and infrastructure behavior to detect unusual patterns, potentially identifying security threats in real-time.

13.Code Review: AI can assist in code reviews by pointing out inconsistencies, potential improvements, or even style guideline violations.

14.Dependency Analysis: AI can predict potential issues with third-party libraries or dependencies, suggesting alternatives or pointing out updates.

15.Anomaly Detection in Systems: For systems in production, AI can monitor logs and behaviors to detect anomalies, potentially predicting and preventing system crashes or failures.

Writing

1.Grammar and Spell Check: Tools like Grammarly and Microsoft’s Editor use AI to offer real-time grammar, punctuation, and style suggestions.

2.Autocompletion: Just as Google predicts your search queries, writing platforms use AI to predict the next word or sentence you might want to write. This can speed up the writing process, especially for common phrases.

3.Content Generation: AI models, like GPT-3 (which you’re currently interacting with), can create coherent and contextually relevant paragraphs, stories, and more. They can be used to draft content, answer questions, write essays, or even create fiction.

4.Style and Tone Analysis: Some tools analyze the tone of a piece to ensure it aligns with the desired sentiment – be it formal, informal, joyful, sad, etc.

5.Plagiarism Detection: AI-driven platforms like Turnitin and Copyscape can compare a piece of text against vast amounts of content online to detect potential plagiarism.

6.Content Summarization: AI can summarize long articles, reports, or documents into shorter versions, capturing the essence of the content.

7.Language Translation: While not exclusively for writing, tools like Google Translate utilize AI to provide translations, helping writers reach global audiences.

8.Content Recommendations: If you’re writing about a specific topic, AI can suggest related topics, articles, or research to include.

9.Chatbots and Script Writing: AI can be used to create conversational agents or chatbots. It can generate potential responses or scripts for interactive scenarios.

10.Book and Movie Script Analysis: AI models can analyze the structure, plot points, and character arcs in books and movie scripts to give feedback or predict audience reception.

11.Assisting Non-Native Speakers: AI can help non-native speakers write more fluently in their second language, making communication more effective.

12.Generating Ideas: AI can help generate writing prompts, blog post ideas, or research topics based on trending topics or datasets.

13.Customization for Audience: AI can be trained to customize content based on the reader’s profile, location, preferences, or reading history.

14.Writing Assistance for the Differently-abled: AI can convert speech to text, helping those who might have difficulty typing. Conversely, it can read out text for those with reading difficulties.

15.Improving Readability: AI can analyze text to determine its reading level and suggest edits to match a desired complexity.

Powers and limitations in writing

As you can see AI is super helpful across the board and when it comes to writing, I am impressed at how well it performs. These lists are a laundry basket of positives so I’m not going to dwell on them. Here instead is where I think AI cannot do a better job than I 🤓

1. When I am writing to elicit action and need context, there are now tools like Microsoft 365 chat to do that (video) to provide that. It does an awesome job. I am blown away by the effectives of providing context. But when t comes to translating that context into meaningful writing that elicits action, it still needs my intervention to truly craft the impact statements that will work. Only I know what truly needs to be done to get the results necessary for success and this is based on my knowledge of the clients business, the relationships and my years of experience.
2. Writing a fun piece of text to use in an Out of Office auto reply is all well and good to raise a laugh and here AI excels. But writing a piece of prose that elicits emotion and human connection, that’s another thing. I think to make a human connection that is personal, something more is needed than AI can provide. I’m sure that AI can write something that captures the spirit of say, a Hemingway, but to write in a bold new style that relates to the sentiments and events of today and creates deep meaning for humans – that I would argue requires a different art form and a human.
3. Ultimately its about voice. I have been writing for years and still don’t think I have found mine. But I am getting better and closer to it the more I practice. It’s something unique to me and my experiences and what makes it special. I’m not saying special enough to make me a bestselling author but I know when I have come close to expressing my voice in a piece of writing when I achieve it. It gives me satisfaction and maybe it resonates with another human being and creates a spark of something. That spark is the glittering bond that binds humanity and cannot be easily replicated.