matt1

Thanks, totally agree about sharing screenshots. I originally started building the site intending people to share text-only examples, but it gets complicated with the formatted responses that ChatGPT generates. And since everyone is sharing screenshots right now anyway, it made sense to just support that for this v1. I do want to add support for sharing text responses so screenshots aren't necessary in all cases. And maybe a Chrome extension could make the process easier, including working with formatted output. Thanks for sharing yours; it could be a good starting point for one build for LearnGPT.

I rolled out a temporary fix which should resolve the image loading issues.

Err... still an issue... working on it.

Sorry about that - the site exceeded my plan's allowable bandwidth on Filestack. I upgraded to a beefier plan so it shouldn't be an issue anymore.

Great feedback, thank you! I just rolled out an update to reposition and increase the size of the buttons to change the image order. The site is using FileStack for the image uploads and as far as I can tell there's no way to tell the order that the images are selected for upload (the callbacks are fired when each image is done uploading, with no details about the order the user chose). Will dig into it more though. Thank you again.

Ah, I understand what you mean now. It's tricky because most people won't care to add a description and requiring one would add friction to sharing. Could do something like having the title be the description if one is present, or falling back to the prompt if not. Or giving users an option to choose. But might be more trouble than it's worth vs always using the prompt for the title.

You're not the first to ask, hah. But no, ChatGPT was not used in the coding of this project :). I likely will add a lot of (human-generated) content around using ChatGPT for coding though. Stay tuned!

Can you elaborate on what you mean by "posting a summary of what the prompt actually results in"? The screenshots + description seem like they take care of both the full ChatGPT response + any summary you might have of it. Can chat over email if you prefer too: help@learngpt.com. Regardless, thanks for checking it out and sharing those novel prompts!

Thanks, glad you like it! If you have any suggestions or run into any issues, drop me a note anytime: help@learngpt.com.

I'll likely add email-based sign up at some point, but it increases the likelihood of spam compared to allowing Google OAuth only, so it made more sense for this v1 to exclude it until I set up more moderation capabilities.

Anyone can sign up and share example prompts, whether it's one they came up with or one they saw online (in which case they can give the source credit).

The submissions are then ranked the exact same way HackerNews articles are ranked on the homepage, following this algorithm: https://medium.com/hacking-and-gonzo/how-hacker-news-ranking... - in short, it's based on how many votes the prompt has received and how long since it was posted.

Funny you should ask this because I just launched https://www.learngpt.com, a site for browsing and sharing interesting ChatGPT examples.

Here's the launch thread, posted several minutes before you asked this: https://news.ycombinator.com/item?id=33923907

You'll find a lot of great examples on the site (and please share more as you find them).

The git log command will generate a comma-separated file which is what the tool expects. The CSV won't include commit messages.

Drop me a note and I'd be happy to delete it: help@preceden.com. Eventually will build a way to do this in-app too.

Creator here. As you realized, it is limited to top 50 contributors and is automatically sorted by descending commit count.

Love the suggestions for tweaking this behavior - added them to my todo list. Thank you for the feedback!

I spent some time last year learning about the math behind neural networks and summarized it in this step by step post that you might find helpful [1]. The post has been on HackerNews in the past and is the top Google result for "backpropagation".

[1] http://mattmazur.com/2015/03/17/a-step-by-step-backpropagati...

I suspect that if you had a hard limit of 45 minutes after which time you could no longer work on your task, you'd adapt and find yourself able to get in the zone much more quickly.

It also becomes easier the more you do it. I've been doing side projects for many years.

At the time of the acquisition the site made money in two ways:

1. Affiliate sales: when you click on a domain and register it through GoDaddy, Namecheap, etc, I get a % of the purchase price.

2. Premium accounts: Lean Domain Search only showed the top 150 results for free. You could pay $79 once for two months of access or $199/year for access to all 5,000 results.

Each was bringing in around $1,000-$1,500/month with the premium accounts growing quickly.

After the acquisition, we made the site free but left the affiliate links in place. The site has grown a lot since then [1] but the affiliate revenue still brings in the same $1,000 - $1,500/month because GoDaddy (the main driver of affiliate revenue) made some changes to its affiliate program that led to lower payouts.

[1] http://mattmazur.com/2015/07/03/lean-domain-search-at-3%C2%B...

Great question - I think the more well-defined your next task is, the easier it's going to be for you to context switch because you don't have to take time to figure out what to work on next. One thing to experiment with is when you're done working for the day, jot down a note of specifically what you want to work on the next time you sit down at your computer. The clearer it is and the less amount of problem solving it requires, the easier it's going to be for you to get back into the flow.

Yep, that's correct: I had my own laptop and commercial internet access through most of the deployment.

Hey all - author and longtime HackerNews user here. It's a nice surprise coming back from an evening away from the computer and finding this at the top of front page. Whatever success I've had with Lean Domain Search is due in large part to what I've learned from this community.

Here's the original HackerNews launch post from 4 years back for context: https://news.ycombinator.com/item?id=3470977.

I'm happy to answer any questions about Lean Domain Search, the deployment, etc.

If you're struggling to find a solid available .com for your startup, I'd encourage you to check out my free service, Lean Domain Search [1]. The site asks you for a keyword then pairs your keyword with 5,000 other words and instantly shows you which are available. For example, if you search for "food" [2], it will pair it with "free" and check "FreeFood.com", "hub" and check "FoodHub.com", and so on, and it also allows you to sort the results by length or popularity of that additional word. With a few searches you should be able to find a great domain, saving you a lot of time and potentially a lot of money.

[1] http://www.leandomainsearch.com

[2] http://www.leandomainsearch.com/search?q=food

I recently spent some time learning how backpropagation works and wound up putting together this step by step tutorial that you might find helpful: http://mattmazur.com/2015/03/17/a-step-by-step-backpropagati...

This is a timely post for me because I also want to study deep learning and have been researching the best way to do that without going back to school and getting a graduate degree in CS.

My tentative plan is to start with Coursera's Neural Network course and implement as many things as possible (probably in Python) along the way. By the end of 2015, I'd like to build a deep learning application from scratch that does something interesting.

While online resources will likely be my primary source of information, I'd love to have a study partner (or partners) to discuss concepts and bounce ideas off of. If OP or anyone else is interested, drop me a note at matthew.h.mazur@gmail.com and we'll figure out the best way to do that.

As a side note, I built a few a-life/AI visualizations in JavaScript over the past year, including a simple neural network which might interest anyone reading this: http://www.emergentmind.com/.

A bit of a belated response, but check out Steven Levy's Artifical Life book:

http://www.amazon.com/Artificial-Life-Frontier-Computers-Bio...

It's slightly dated, but I thought it was a really good introduction.

Hey David, I run a similar service called Lean Domain Search [1] and was excited to check out your service.

A few ideas:

- If you're willing to consider limiting the TLDs to just .coms, you can then use the .com zone file to check the availability of domains much much faster than methods. Using this technique Lean Domain Search is able to check the availability of 5,000 domains in about a second. I wrote about how to get access to the VeriSign zone file here [2]. Parsing it so your app can access it in real time is tricky, but I can go into more detail over email if you want to explore that route.

- One thing to think about is whether figuring out the search volume is that important for your users. If it's a challenge getting that data from SEMrush or if it's slow or if it just clutters up the interface, consider dropping it and focusing on the quality of the suggestions. More information is not necessarily better for users (it depends what problem you are solving and for whom).

- If you want to a quick list of prefixes and suffixes to pair the user's search term with, I previously analyzed all registered .com's to figure out what the most common ones are. That list is available here if you're interested [3].

- I wound recommend identifying a specific pain point that a specific group of users have that isn't addressed well by existing tools and focus all of your efforts on solving that. The more focused you make your app, this better.

Props on getting users to tweet about the app to show more suggestions. I used to do that with Lean Domain Search and it works really, really well.

I'm happy to chat over email if you'd like. See my profile for contact details.

Best of luck with this -- congrats for launching!

[1] http://www.leandomainsearch.com

[2] http://www.leandomainsearch.com/blog/16-how-to-get-access-to...

[3] http://www.leandomainsearch.com/top-domain-name-prefixes-and...

For sure -- that is one of the medium term projects I want to work on. We'll see how it goes in the browser though...

If you're interested in emergence and artificial life, I've been working on creating in JavaScript a lot of the work that has been done by others in the past as well as some original projects:

http://www.emergentmind.com/

The last project (The Evolution of Color) has similarities to the color visualizations described in this post: it uses genetic algorithms to evolve a population of colors towards a fittest color as determined by the closeness of the bits that comprise the colors.

I have a few other projects planned so if you're into this stuff you're welcome to sign up for the mailing list to get notified when they are launched:

http://www.emergentmind.com/newsletter

I'm always open to ideas for new projects so if you have anything you'd like to see implemented or want to work on something together feel free to shoot me a note.

Ah yes. Here's what's happening:

When the biomorph gets too big for its canvas, I automatically reduce its branching depth until it's not too large. If you keep clicking on children that make it evolve towards larger entities, eventually it will shrink and look something like the biomorph in your screenshot.

Someone suggested that rather than automatically reduce its size, I should just make them unclickable (ie, you can only choose a child that fits within the canvas).

What do you think? Any other ideas?

Edit: I just updated the implementation so that instead of automatically reducing the child biomorph's size if it's too large, it simply becomes unclickable (you'll also get an notice if you try to click it). Should be a bit better -- thanks for the nudge!

Thank you for the kind words.

I have other JavaScript-based emergence and artificial life projects planned so if you liked these, you should sign up for the mailing list to get notified when new projects are released: http://www.emergentmind.com/newsletter

If anyone reading this finds these topics fascinating as well I'm always down for a chat: matthew.h.mazur@gmail.com or @mhmazur - cheers!