To set the stage, I’d like to share a thought-provoking quote from Lex Friedman about the future of AGI (Artificial General Intelligence) from the podcast intro with Sam Altman:

…Soon, nobody knows when, but many believe that it can happen within our lifetime. The Collective intelligence of the human species begins to pale in comparison by many orders of magnitude to the general superintelligence in the AI systems w build and deploy at scale.

This is both exciting and terrifying. It is exciting because of innumerable applications we know and don’t yet know that will empower humans to create to flourish to escape the spread of poverty and suffering that exists in the world today and succeed in this old all too human pursuit of happiness. It is terrifying because of the power that super-intelligent AGI wields to destroy human civilization, intentionally or unintentionally. The power to suffocate the human spirit in the totalitarian way of George Orwell’s 1984 or the pleasure fuelled mass hysteria of a brave new world where, as Huxley saw it, people come to love their oppression to adore the technologies that undo their capacities to think.

The pace of technological advancement is truly unprecedented. I’ve never seen anything like this before in my life. I don’t consider myself a flashy AI evangelist. Still, given all of the OpenAi’s and myriad other AI startup breakthroughs, I wouldn’t be surprised if we witnessed things in the coming year that seemed unimaginable in fantasy novels.

Like Lex Friedman, I’m excited and terrified as a person and professional by all the latest advancements and the possibilities that lie ahead, imaginable and unimaginable. And I keep asking myself a question. What can I do to ride this AI wave or at least not be entirely dragged by it?

A person typing a prompt to create a product

The future of prompt to product

Overall, I’m pretty optimistic about the potential of LLMs and AI. Seeing Greg Brockman demonstrate what GPT-4 could do on a recent developer livestream — like turning a simple sketch into a fully functional app — really got me thinking about all the exciting possibilities that lie ahead.

Imagine you’re a product manager, designer, or founder with an idea to test:

• You write a simple PRD with a basic product description, user stories, and hypothesis and then submit it to an AI tool. Let’s call it ‘Prompt to Product.’
• With just a click of a button, the tool generates a working app and deploys it to your local host server.
• You test the app and identify issues, which you describe to the tool as you would to a developer, and continue to dialogue with the tool until satisfied.
• Then you ask the tool to deploy the app to the testing environment.
• When your app is online, you use another tool called ‘Prompt to Test’ to conduct unmoderated user testing with external users. Just upload a list of emails and pick the hypotheses to test, and ‘Prompt to Test’ handles the rest of the testing process.
• Shortly, ‘Prompt to Test’ provides you with general observations and suggestions for improvement, which you can accept or modify.
• You might involve a designer and an engineer to make your product even better and optimized for scale.
• The designer can tweak the UX using text prompts and image sketches.
• The engineer can look at the code to fix potential database issues and add privacy improvements, for example, all with a couple of prompts and little to no coding interventions.
• After running through some tests, your product is ready for release to the world.
• And the best part? You accomplished all of this in just one week with a couple of back-and-forth prompts!

And that’s just what I was able to imagine. Just think of all the things we don’t know are technically possible yet that will emerge in the coming months and years.

The future is looking bright for product people, founders, and creatives. With the advancements in technology, you won’t even need someone to write code for you to bring your product to market. The “why” behind your product will become even more important, if not paramount, than the “how.”

Of course, the downside is that there will inevitably be a lot of crappy products and copycats. Still, on the bright side, it will also unleash human creativity, resulting in diamonds shining through the barrels of mud.

Current reality

However, let’s take a closer look at these tools now. It’s still necessary to have coding skills and a certain level of familiarity with the technology or framework you’re working with to ensure your app works as intended.

But the process has become significantly more accessible, as you mostly need to understand the general principles and approaches of how to make digital products and a lot of patience to debug the code written by tools like ChatGPT. But with each consecutive prompt, you can create a viable prototype to test your idea with the world.

Moreover, here are a few examples of how you can utilize these tools:

And you can find dozens of articles online that teach you how to create better prompts for coding with ChatGPT, such as the ones linked here and here.

Oh, I almost forgot to mention that tools like Copilot X take this process to the next level by eliminating the need to switch between your IDE and browser.

How to ride the wave

When I began writing this article, the main focus was on the question, ‘Should product people code?’ but as I dug deeper, it led me down a rabbit hole of envisioning the future of AI.

So, the answer to whether product people should code is a resounding “yes.” And I want to emphasize that they should not only know how to code but also be able to code themselves. Because until there are tools like ‘Prompt to product,’ you need to fix some minor things yourself to get the prototype running.

Indeed, as these ‘prompt-to-product’ solutions become more sophisticated, product people will require fewer and fewer coding skills. However, gaining foundational knowledge will help you better understand the systems. And in a future where anyone can create a product with a single prompt, having this knowledge may provide a competitive advantage.

Going back to the question I posed at the beginning of this article — “What can I do to ride this AI wave or at least not be entirely dragged by it?” — I have a solution, at least for the beginning of the 2023 version of myself. I hope you’ll find it helpful too.

My plan as a product manager who has transitioned from product design and has an engineering background is as follows:

• Refresh my coding skills.
• Build small practical tools with the help of LLMs or so-called ‘front-ends’ for LLMs.
• Improve my prompt design skills, especially for professional needs.
• Gain a better understanding of how LLMs work.
• Go deeper into recent AI research papers beyond those trends that lie on the surface (this is tricky).