Oh yeah, I agree with that worst offender - features that don't exist that it quotes so confidently.
Especially when libraries get updated constantly - I ran into cases where it wrote functions based on previous versions of the library, where the function names and signature had already changed since.
I haven't tried yet, but I bet if I used a version that has web browsing and can read the latest documentation, it would do a good job and save me having to read the documentation myself and try to work it out.
I only have a little coding background, and ChatGPT didn't make me an expert. But the collaborative Human+AI process did allow me to complete a project end-to-end, including figuring out where to host it and how to do that.
I found that it helped me with 6 "superpowers":
1. Choosing between options (e.g., AWS vs. GCP vs. Zapier)
2. Walk me through it (e.g., how to set up a Firestore database)
3. Text-to-code (including simple nuisance calculations and code-to-code changes)
4. Help me out! (i.e., fixing broken code based on error messages)
5. Teach me (e.g., learning the difference between let, const, var, etc...)
6. Check my code (e.g., it caught errors before I even ran the code)
Check out the post for more details if you'd like!
There's another post on building a website from scratch where I also tried Replit's Ghostwriter. Yes, I faced a lot of frustrations in the process, but going from "I can try struggle through this on my own" to "I actually have some help here that's always available and usually right" is amazing IMO.
I hadn't coded in ages. I started a hobby project to see whether I could remember enough to "set up a cute chatbot over email" with ChatGPT's API.
At some point, ChatGPT asked me what added value users would get from the project. That was an interesting question!
It was also an excellent example of how this project became more of a collaboration instead of a one-way street of asking ChatGPT to do stuff. The back-and-forth brings in added value beyond each of our contributions.
So we (the AI and I) decided to document this collaborative journey on the project. So far, we've worked on:
1. Planning the project and pieces of it: excellent for frameworks, ideas, and feedback.
2. Coding + tech setup: we wrote about six superpowers, from choosing between options (like AWS vs. GCP vs. Zapier) to learning code best practices and more.
3. Building a website with ChatGPT and other AI tools in a day.
To try it out, install then imagine you're planning a trip to somewhere - e.g., type "Things to do in Paris" into Google then check how it organizes the information for you. Feedback most welcome!
Loved this post - for any others here who share similar views, thought I'd share a few thoughts from our experience working on this.
I'm co-founder at https://www.wanderant.com. We started off with similar frustrations on planning trips. All the information you could ever want is online - and for free - so we should better off than where we were when you'd just buy a Lonely Planet book. But that's not the case, I think it's actually gotten harder - partly because there's so much info, and so many options.
We're a year+ into it, and have been iterating multiple times on an MVP. It's a big problem to solve, so 'minimal' is a bit high - plus it's a different minimal for every traveler you meet (e.g. comment here on "an option for children" which is absolutely critical for some).
Here are some things we've learned along the way:
* Top 3 things people need / want that haven't been solved:
- Local expertise - esp when traveling abroad - what is the local yelp / gothamist / etc?
- Help with logistics - mostly around how to get from A to B - Rome2Rio is a great help in that regard, still takes a lot of work for planning an entire trip though
- A way to manage a plan in one place - instead of email /bookmarks / excel
* Some of the challenges we've encountered:
a- Making a product that's easy to use - If you're only solving one step (e.g., online booking), it's easier to create a workflow that makes sense. I love examples like how hipmunk have simplified UX for booking flights. We are building towards a workflow that lets you keep your entire plan in one place - ideas, map, notes, reservations,... and finding a UX to make that easy is a lot of work.
b- Getting to a product that really adds value to planning- We want a tool that simplifies the process of planning. However, we clearly can't start with a product that does everything you need for a trip. Even in the near future, our product will continue to be an incomplete solution. So paradoxically, by being an additive tool in the process, we've successfully worked against making it simpler :)
We have to make up for that by really adding a lot of value and saving time fr the user- and getting to that point is also a lot of work
c- Marketing is hard - others have mentioned the challenge of facing giants in the space- I'd like to offer another piece- which is reaching trip planners at the right time. Only a small % of people we reach are in the process of planning a trip. Then if they are, we most likely don't yet offer very rich content for where they're going - because currently our higher-quality content is still nascent. So that makes finding the right users hard. Any help here is appreciated btw - if you know someone who could use this send them along :)
* Why we're continuing
- Times like this, when we see someone get passionate about how much they wish this existed or how they can't believe it hasn't been done before. We feel the same :)
- While it's a hard problem to solve, and there's a lot of challenges we already know of and more we don't yet - it is a really interesting problem to think about and try to crack.
- More than all, we love to travel to new places and explore the world. If we can share that experience and make it easier for a few others, we'd be ever so pleased.
Especially when libraries get updated constantly - I ran into cases where it wrote functions based on previous versions of the library, where the function names and signature had already changed since.
I haven't tried yet, but I bet if I used a version that has web browsing and can read the latest documentation, it would do a good job and save me having to read the documentation myself and try to work it out.