Flow States and AI Agents
Keeping focus and choosing the right problem
A couple of reflections on recent podcasts:
Maintaining a flow state with AI
Author Seth Stephens-Davidowitz described in a podcast how ChatGPT empowered him to write a book in only 30 days, thanks to the uninterrupted flow states he could achieve when generating data visualizations:
The difference for me…is night and day in how much more joyful, free, and fun it is to use ChatGPT. Editing and adjusting your code—half the time you still make a mistake, wondering where's the bug in the code. Going from looking up the code, writing the code, looking for your bug in the code, adjusting the code, to ChatGPT doing it instantaneously—even if that's just 20 minutes to one minute—is changing something that's 20 minutes, then a 40-minute break where I need to recharge myself, to one minute…That is like night and day in the research process.
This is an incredible benefit of AI. The ability to maintain a state of flow is crucial when focused on working through a challenging project or problem. AI interfaces like ChatGPT are so quick and effective at providing suggestions or solutions, that we no longer have to lose focus by turning to Google and hunting down the right link/page/paragraph to get the answers we are looking for. AI helps us avoid the struggle known as “yak shaving” — the seemingly endless series of small tasks that have to be completed before the next step in a project can move forward.
What problem are AI agents trying to solve?
A common use case I’ve heard for AI agents involves booking trips, for example some variation of what is described here in this podcast:
Let's say you're like, "Hey agent…I want to go to New York tomorrow," and it knows every preference—my seating preference, whatever. It would come back and be like, "Okay, I'm going to ping Wheels Up, I'm gonna ping Kayak, I'm gonna ping everyone and say, 'Hey, Sam wants to go to New York, what do you got for me?'" and it can, in parallel, negotiate with all of them.
But how common is this kind of scenario? People who travel are often just visiting friends or family. And for those traveling for enjoyment or business, there is probably an element of fun and autonomy in researching things to do.
Of course, travel companies like Expedia have already integrated AI into travel planning, and that can be useful for providing suggestions. But if planning business trips is the number one use case that tech founders and VCs can come up with, that reflects a lack of imagination.
Many of the individuals starting and funding tech companies have very different lifestyles than the average person. Companies like Remitly, whose mission is to help immigrants send remittances back to their families, could have really interesting applications of AI agents. One reason why a lot of successful companies have immigrant founders is not just their work ethic, but likely also their worldview, which gives them the perspective to better understand markets, customers, and problems that need to be solved.