The Rise of AI - What it Means for Writing and Writers

Generative artificial intelligence (AI) has the potential to revolutionize the field of writing and journalism. With the ability to generate human-like text, generative AI has the potential to assist writers in the creation process, enabling them to produce more content in a shorter amount of time. However, there are also concerns about the impact of this technology on the future of writing and the role of human writers. In this blog, we will explore the potential benefits and drawbacks of generative AI for writers and the future of writing.

I didn't write that paragraph, ChatGPT did in about 5 seconds based on a prompt I gave it.

ChatGPT is a sophisticated chat bot developed by OpenAI , a private technology firm focused on artificial intelligence (AI) research. The company was founded in December 2015 and its stated mission is to "advance digital intelligence in the way that is most likely to benefit humanity as a whole". The whole tech world has been buzzing about ChatGPT's capabilities, which seem almost limitless ranging from creating detailed workout programs to sonnets written in the style of Shakespeare, all from simple prompts fed to it like the one I gave it to start this newsletter.

What is generative AI technology?

Tools like ChatGPT fall into a category of technology folks are calling, 'generative AI', which refers to a type of artificial intelligence that has the ability to generate content based on its training from massive datasets. For a deeper dive into the tech and the people behind it, I recommend listening to an episode of the podcast, Hard Fork, called "Can ChatGPT Make This Podcast?". I'll spend some time in this piece going over what I believe are some important implications to think about as generative technology transforms the nature of work and grant writing, in particular.

To find our bearings, I'll begin with a quote from James Currier , General Partner at venture capital firm, NFX , who writes, "Until now, software couldn’t solve the zero to one problem because it worked FOR us. Generative Tech will work WITH us from the beginning of any project." (article) Currier's writing on generative tech has helped crystalize many concepts for me and recently led me to get out of bed as I was falling asleep one night to write the following on my whiteboard:

We're used to being the only "being" when we use software, so we focus on what we get from it. We're paying the most attention to what generative tech "generates". A profound shift taking place is we are no longer the only being, the only perspective. Just as deserving of awe is the AI's ability to read and understand. Reading is the precursor and origin of its amazing generative abilities. Yes, it can write, but more fundamentally, it understands.?

Generative technology represents a major step toward genuine intelligence in software, not only because it can create impressive outputs, but also because of how well it understands what we're asking it to do. When Currier notes we will go from software working for us to working with us, I hear this as an acknowledgement of a fundamental shift in our relationship to software, a flattening of hierarchy, and an acknowledgement of its greatly increased ability to understand us.

There are some wonderful possibilities that arise from technology that is so capable, and also very legitimate reasons to proceed with thoughtfulness and care as we learn to work with generative technology.

What does writing with AI feel like?

I've been writing with some AI assistance for a few months now and the way I've come to describe my experience is to analogize it like this:

Writing without AI is like walking, while writing with AI is like riding a bicycle.

I can't tell my bike where I want to go and expect it to take me there. I have to pedal, balance, and steer my way to the destination. Similarly, it hasn't been my experience that I can give the AI a prompt and receive a response that doesn't require any revision.

Quite often, what I get in response to my prompt is unusable. Yet even when this happens, I'm not discouraged, nor do I feel like I've wasted my time asking the AI to try to help me. It feels similar to leaning over to ask a colleague a question and they happen not to have the answer I'm looking for. No harm done.

What are some advantages of writing with AI assistance?

By riding a bike, I can travel distances with much greater speed using far less energy than I can by walking. I've had this same feeling writing with AI assistance. AI has helped me to write bullet point outlines to help me structure a piece, it's helped me to add additional detail when my own knowledge is lacking, and to quickly rewrite sections to be more concise, or more clearly written. In many cases, just like a riding a bike, I can reach the same word count more quickly and with less mental energy.

Here are some of the ways I think AI can be helpful to grant writers:

• Summarizing large amounts of research material
• Creating outlines
• Simplifying complex concepts
• Improving grammar and structure
• Moving through writer's block
• Adding more depth of knowledge

It's important to note how fast this technology is improving, and I'll say more on this later in this article. These are some of the features that feel comfortable and effective to use now, and I fully expect this list to be out-of-date within a few months.

What to be cautious of when writing with AI assistance?

Continuing with the bike analogy, imagine you've just crested a hill and you're beginning to coast down the other side. The view is gorgeous, warm air feels great rushing by, and you take your focus away from steering the bike to film a quick video to show your friends, meanwhile you're headed right for an uneven chunk of pavement sticking up in your path.

Luckily, you avoid serious injury and, hopefully, the bicycle analogy also helps to illustrate potential hazards of working with generative AI. Coasting along, one could easily overlook opportunities to shape a piece of writing into something truly extraordinary, or miss subtle conceptual misalignments that add up to a weak or lackluster overall narrative. In other words, it could make us lazy writers and dull our skills.

A lot of students are already using ChatGPT to do their homework. For example, you could tell it, "Write me a book report about the book, Of Mice and Men, discussing the relationship between the main characters, written at an 8th grade level." And this is what you get:

I'm reminded of a quote from the book, Bird by Bird, by Anne Lamott , in which she says,

"Almost all good writing begins with terrible first efforts. You need to start somewhere. Start by getting something—anything—down on paper."

What happens if we no longer have terrible first efforts? What are we losing? I think one of the most important things we must pay attention to as we learn to work with AI is to protect and value our creative processes. Writer's block is a common experience in the writer's creative process and it can be really painful. AI allows us to bypass this pain. When should we take the shortcut, and when should we stay on the scenic route? Those who navigate these questions well will be the masters of working with AI.

Here are some habits I'm building as I learn to write with AI:

• Consciously protecting, nurturing, and valuing my creative process
• Paying attention to my narrative voice and how it changes with AI
• Regularly reviewing AI-written content
• Tracking the quantity of my writing
• Sharing my experience with others using AI writing tools

What should I do if I want to learn more about writing with AI?

First, take a breath and give yourself credit for being open to exploring such profoundly powerful new technology. The emergence and ascendance of generative AI exemplifies the accelerating speed of change in the world.

Once you're ready to dip your toe in, here's how I would begin:

1. Create an account to try ChatGPT for free using this link
2. Once you're in, ask ChatGPT, "What is generative AI?"
3. Once it answers you, tell it, "Explain all that so a five year old can understand."
4. Keep going and see where your conversation takes you

I also recommend listening to the podcast, Hard Fork, hosted by Kevin Roose and Casey Newton , which covers big stories from the frontiers of new technology. They have a few recent episodes about generative AI, and I also enjoy how they make learning about all kinds of different technologies fun and relatable.

What else should I know about AI technology?

Understand how quickly this technology will improve.

The amount of data available to train AI models has increased dramatically in recent years due to the proliferation of internet-connected devices, like smartphones, which has allowed people to create enormous amounts of content for AI's to learn from. Now, with the help of AI, content can be created even faster, at greater volume, and it can go back into models to train even more powerful AIs.

The world's largest technology companies like Microsoft and Google, along with tens of billions of dollars of investment, are putting enormous amounts of computing power and money behind AI's advancement, which will also drive the speed of its improvement.

We're going to feel AI more than any other kind of software.

AI will be a presence in our lives. We'll perceive it learning and becoming more intelligent, more perceptive, more intuitive, and more creative just as we do with children. We're going to feel delighted and intimidated, grateful and disdainful, seen and surveilled. We're going to feel AI, a lot.

The emotionality of this new era will be just as novel as the tech itself. Just watching the range of emotions elicited from my friends and loved ones trying ChatGPT gives me a glimpse of how society may react when more fully fledged AI applications are out in the world.

Conclusion

I thought about asking ChatGPT to write the conclusion for me and I bet it could. Instead, I thought I'd ask it one of the many new questions that arose for me over the course of writing this piece:

In the exchange above, ChatGPT again demonstrated profound intelligence by quickly and precisely defining 'wisdom' for me when I asked for it. Next, when I asked it if it is, itself, wise, ChatGPT's answer was complex and thought provoking. It seems to consider 'living', 'personal experiences', and 'emotions' as key ingredients of wisdom and then describes its own abilities and limitations relative to human wisdom. A pretty wise answer, in many respects.

I find myself hopeful and intrigued by the possibility that artificial intelligence could someday give rise to real wisdom, from an artificial source. A wise AI working throughout the world to help us all be better versions of ourselves would be a powerfully good force. I think we have to keep imagining the future we want, sharing and shaping that vision with each other, and then taking collective action to ensure it becomes reality.

Thanks for reading this first edition of The Process. While writing this post, I got feedback to do another post on the implications of AI on grant writing, in particular, so look out for that article in the weeks to come. Please feel free to share comments or you can email me at [email protected]

Philip Deng is the CEO of Grantable , a company building AI-powered grant writing software to empower mission-driven organizations to access grant funding they deserve.