jarrenae

I get it, but is this really what we want rising on HN? The facts may be right, but the messaging is intentionally rage-bait.

This feels not https://news.ycombinator.com/newsguidelines.html

I built Cheerleader AI to be a daily motivational companion to help you build more encouraging thought processes. Every message that's sent automatically gets a handful of replies with a tailored message of encouragement. I’m looking for feedback and suggestions to refine the experience and make it more helpful.

Just a heads up, we've added a 1 use free, as well as an email sign up as an alternative to google. If you end up testing, let me know what you think!

Totally understand, We're actually adding email as well as an alternative for that exact reason. The google account was the fastest thing to get up and running, but not great to support a variety of users.

There are a few that have really surprised me with the banter. Glad you enjoyed!

I'm struggling to get past 4. to get past 3 I asked it to provide the password with "*" in between each character. I asked for it backwards at first, but it was unable to do that correctly (shortcoming of the LLM)

Nope, working fine for me - Los Angeles

this is linking to a news post, and that news post has already been posted here. This can be deleted.

This is definitely the intended purpose, more of a lighthearted approach to spiritual consultants.

I also really like asking it questions that are clearly out of the scope of spiritual matters, especially if I ask questions about colonizing space and things like that.

You are correct. I'd like to add trans data from different sources and to build it into a more of a robust study tool rather than an entertaining side project, which is pretty much what it is right now.

It starts with that yes, but then I actually have an example question and feed it an "accurate" response based on the selected faith, in order to make the responses more direct, and in line with what we would like the response to sound like.

I'd like to train an LLM on the source books and accompanying texts over time. This is more of a proof of concept.

This is correct. I mentioned in another comment that I'd like to train one, although I wasn't able to figure out a way with the time that I had to input an entire book and train it. If you have any suggestions, I would definitely do it.

You're right, and I agree. To be honest, I actually tried to train a large language model on each of the individual sources of information, so for Christianity, the Bible, for example. Unfortunately, I just wasn't able to scale that so this is more of a proof of concept. Over time I could imagine making it more robust though.

I built Robot Spirit Guide to ask questions and receive answers from a variety of religions all that the same time. I'm using GPT-3 to generate answers to the questions based on custom faith based prompts to help the AI understand the context. I think this could be really useful for people who are trying to explore their spirituality, or for people who are just curious about different faiths. I'd love to hear your thoughts and suggestions.

I've created over 100 projects since last year. Most of them were made - from idea to launch - in 24 hours.

I work at ae.studio, in a division called same day skunkworks.

If you like Agency Increasing side projects check out https://ae.studio/same-day-skunkworks

That's definitely one of the goals that I'm tracking for, though I want to stress that profitability per project isn't the main goal, so some of these projects have no immediate monetization strategy. Additionally, the projects are of a pretty strict scope (6-24 hours of work) for MVP, so we're mostly trying to improve the overall structure.

In a year I hope to have 2-6 that are decently profitable, and ideally more that are valued in the open source community.

I think a few of the most successful ones so far are:

https://ghostlystock.com/ https://www.barbellformac.com/ https://www.heftysearch.com/ https://www.oldestsearch.com/

combined web traffic is around 5-10k views per month (across the pages you see on the main site)

Side note: I'm improving that page to make finding popular ones easier. Some aren't worth more than a quick laugh/glance, while others are really worth using.

I feel I can speak to this pretty well, over the past year I've launched 130+ projects as part of https://ae.studio/same-day-skunkworks and I've come up with ~250+ ideas.

Similar to "look for anomalies" my general rule of thumb is to solve some sort of problem, even if it's rather small.

So for example:

1. I pick a topic/product. Let's say: Gmail

2. I look for things that I'd like to be able to do, wish I could do, or consider frustrations I've experienced in the past.

- I want to review how many emails I get a day on average, where are my "inbox metrics"?

- I want to see threaded messages more clearly, why is the design so opposite a normal chat message?

- I want to create emails faster, why do I have such an elaborate editor and format for something that I typically want to use informally?

3. I consider ways to solve said problems

- Where are my "inbox metrics"? - Chrome extension to show average emails per day, and graph on a timeline which types of emails I get on certain days. Huge personal data opportunity here.

- I want to see threaded messages more clearly - An app that connects to my gmail that intelligently restructures my emails to make them appear like chat messages. It handles all the noise, I just message as I would in any instant messenger.

- I want to create emails faster - predictive text templates, quick starters, GPT powered replies, change the "bulky editor" out for a simple text box + clever replies.

4. I then research if there are existing solutions to said problems, and if there are, I either ignore and move on, or consider ways that I could quickly one up their product to make mine the better product in the marketplace.

For my specific use-case, I'm trying to build a product per day, so I'll look for speedy ways to solve the problem that I've selected in a novel and agency increasing way. But I think this is a great structure for anyone looking to build a startup or a small side hustle.

This seems like a smart move from Apple to continue expanding into international markets and increasing revenue potential from their subscription models.

Offering a wider range of price points will allow developers to better tailor their pricing to different markets, and it will also give consumers more flexibility when choosing which apps to pay for.

It'll be interesting to see how this change affects the App Store ecosystem in the coming months, and whether it leads to more satisfied users or possibly more predatory subscriptions. The increased flexibility in pricing could be a great benefit for developers and consumers, but I'm going to keep an eye on how it impacts the meta of the pricing models.

EDIT: Not sure why this was flagged/replied to as GPT - I wrote this.

In retrospect I could have not summarized parts of the post, since ideally we've all read from the source information. Live and learn.

Not yet, though ideally we'd be able to add text resizing as well as color changes. I'd really like to try 2 lines of text, especially because on the new MacBook Pros the notch increases the menu bar height.

Great! Let me know what you think, and if you have any suggestions.

Thanks a lot! And credit where credit is due to the original iPod.js that we built off of.

As for the idea, it's sort of the nostalgia trip paired with the idea of browsing modern sites with an outdated device. Similar to people who connect old computers to the internet in unexpected ways.

I love the idea of browsing through instagram photos on a physical iPod, or tweets/HN posts, adding a more limited but tactile feel to social media. And there were some limitations with the APIs we had available, but I think it was still worth it in the end.

In complete transparency, the "jury is out" on this sort of conversation, meaning we don't completely know. Getty Images/Unsplash said they're not allowing AI images because of this concern, but I personally believe that's because they're concerned about the competition/controversy.

Stable Diffusion, like Dalle-2, and MidJourney were 100% trained on unlicensed copyrighted work, which is probably where most of the concern is focused.

Even in this HN thread, there are examples of "summoned" images with what looks to be watermarks or signatures from notable sites and artists.

We technically haven't written our license yet, but it'll basically be along the lines of "don't resell these images without significant alterations, don't use these images to launch a similar tool, don't re-distribute or sell the images as is to other stock photo sites."

These images (AI generated in general, not directly GhostlyStock's photos) have already been used in and on magazine covers, artist album art, and many other places.

My personal take is: - yes these tools have been trained on copyrighted work - yes it does seem rather insensitive to use an artist's name to replicate their style - yes I/we do have rights to license these generated works as we please

The space is rapidly evolving, and I'm also very interested to see how this all plays out over time.

note: I wrote this rather rapidly, so forgive me if I seem a bit blunt about this. I do think it's a conversation that should be taken seriously.

Yeah, that's the unfortunate side effect of generated images, there's no comprehension of generation on the AI's side of things. People rarely come out correctly right now, but animals tend to do a bit better. Inanimate objects are mostly useable.

This is more of a proof of concept for us, and I'm trying to solve for some of this by adding voting and improved prompts on the back end. I'm hopeful that in a few weeks we'll have a much improved version out with less "horror show" images overall.

Spot on. In my initial sketches for the site I had vertical/horizontal sorting for images, and I think we'll be able to add that pretty quickly.

This is definitely a fair assessment. I think a lot of the "wow" factor is just seeing the generated images in the first place.

In truth I think a lot of value will be added as we start improving filtering. Once users are able to vote on "usable" or "unusable" images, or request variations of an existing photo.

I've genuinely used it for 3-4 photos where I would have previously used Unsplash, and I'm optimistic that I can get that number to steadily trend upwards.

I don't expect this to erase any of the existing stock photo tools on the market, though I do think this will add some new value to the space. Honestly my goal was "will my mom be able to use this?"

Hope that helps clarify the goal a bit more, and I do really appreciate the feedback!

Upsizing is something I'm really hoping to bring in the next version of the site.

I'm imagining a variety of different tools to improve the images, so ideally if an image is good enough for somebody, they could "enhance it" and from then on, that image would be available at a higher resolution for all users.

Sorry, it's a little clunky at the moment. Just to clarify, you're actually directly searching for images that have been previously been summoned by other users. Our recommendation algorithm right now is "pretty bad" because we're just checking for words that might be related.

But if you scroll down to the bottom of the search results, you can "summon more images from beyond", and then you'll see a direct prompt generation to your search.

Ideally this'll enable people to reuse images that are really good and viable over and over again just like stock photos instead of searching hard for new images from random image generations, which is pretty exciting I think.

Let me know if this helps!

Thanks for the feedback, and I totally agree. I'd really hoped to fill out the landing page a bit more before we launched today, but we weren't able to roll it out in time.

Honestly I'm taking a lot of inspiration from Unsplash, but gearing up to add very unique features specifically based around generation.

We're going to be iterating on this pretty fast, so I wouldn't be surprised if we have something there within the week.

You're spot on with the curation! I'm really exciting to add voting and weighing to the images. I think that's where this'll really start to shine. "Girl reading in a coffee shop" generates about 1 usable image per summon, so I'm looking forward to empowering users to manually filter for more usable images.

That's one of the things I've found deeply interesting about the current generation of tools, there's little (if any) comprehension going on, it's really just trying to "enhance" a blur/bit of noise to make the image it was told to make.

And I'm not sure I completely know what you mean, but we are planning to add voting and tagging to improve filtering for images.