ChuckL

Inspiration

(In 1984-esque dystopian radio announcer voice) Live, ~~Laugh~~, Love, ChuckL.

As we continue to move into the Information Age, there is no more need for life. Get rid of those flesh-bags and simply integrate your consciousnesses through our Live© hosting servers and experience anything and everything you could have ever possibly imagined!

Finding a soulmate is so 21st century! Instead, leave it to our in-house Love© algorithm to simulate making the genuine connections that your former anti-social self never could!

And last but not least, fear not, weary travelers of the digital landscape, for we offer respite from the endless quest for humor amidst the chaotic abyss of online forums! Behold ChuckL! Inspired by the dating-apps of yester-year, ChuckL is a beacon of light in the tumultuous sea of memes. No longer must you traverse the vast expanse of Reddit or tirelessly scour Facebook groups in search of comedic enlightenment. With ChuckL, your journey to comedic nirvana begins with a simple swipe.

ChuckL. Why Laugh when you can ChuckL.

What it does

This app is like Tinder but for memes. The user will sign-up/login to create an account and memes will be presented to them. As the user swipes left or right on the memes, their feed will be further fine-tuned using AI to predict content that the user will find funny. Therefore, as the user continues to swipe left and right through memes, their feed will eventually be filled with memes which match their preferences. The users will create a unique login which will track their preferences between sessions.

Who needs to painstakingly scroll through Reddit, iFunny or Facebook to find communities that may or may not share your sense of humor when you can have those niche memes come right to you!

How we built it

For the hosting service we an AWS EC2 instance and our web service is Flask and we created our application using pure HTML, CSS and JavaScript. The underlying scripts to access the images are written in Python, and to create our image recommendation system we used the Scikit Surprise library. Our application was hosted using a GoDaddy domain, and we have all of our files in a GitHub repository. The account creation system for the website is hosted by Auth0.

Challenges we ran into

Initially we had many problems related to web hosting. We tried to host using CloudFlare, but that only allowed for static web pages and didn’t offer the DNS control that we needed. We then tried to switch to Django hosting, however that would not display any of our content, so we switched to Flask instead. It took us a while to figure out how to get the custom domain from GoDaddy/Porkbun onto Flask (our hosting server), since trying to edit the website using the built in features prompted us to use their hosting platform (which was paid, and would have required us to restart from scratch).
Most of us were also unfamiliar with the many tools we ended up implementing, so reading the documentation and trying to understand all of the programs took up a large portion of our time.
We also had a lot of problems when it came to image curation. Initially we wanted to have our program web scrape the metadata from memes on platforms like Reddit and then use AI to recognize which tags are in common between the images that the user swipes right on, however, their recent Reddit API changes made this unfeasible, and other sites such as iFunny have also followed suit.

What we learned

We faced and eventually overcame the various issues related to scaling. For example, our application worked well locally, however when we tried to implement the same solution into our actual web-hosted project environment, nothing seemed to work. Because of this, we had to troubleshoot and re-build our basic idea from the ground-up with the scaling in mind.
We were able to gain valuable experience with various back-end development tools such as Flask and Scikit Surprise, and we were able to deepen our understanding of front-end programs such as HTML, CSS and JavaScript.
We also learned about various tools such as Auth0 through the workshops, which we attempted to implement into our program.

What's next for ChuckL

(intercepted communication from internal ChuckL communications)

The investors have really been on our behinds this last quarter and have been demanding to see proof of our AI content curation system. Currently pesky platforms such as Reddit and iFunny have measures in place to prevent us web scraping and collecting the meta data on their images, so we will need to secure funding to pay for their custom APIs. But of course, it isn't over promising and underdelivering if we do eventually deliver!!!

Unfortunately this means that we will be enforcing mandatory un-paid overtime for our friends at the Machine Learning Engineering department, because if you won't do it, there's 20 more interns ready to jump at the chance to do it instead. Seems everyone wants to work at a sketchy, pop-up venture capital funded start-up these days.

And of course, here at ChuckL LLC ©, we believe that humor is the key to unlocking a person's psyche: unraveling their greatest hopes and darkest desires. Desires such as the products you didn't even know you wanted! Of course, as is industry standard, we will then take that information and sell it to the highest bidders. As part of this initiative, we will also implement an interactive comment system, since as everyone knows the comments are always funnier than the actual meme anyway. This also adds another dimension of profit (user data) which we can harvest, with us now being able to track the user’s mannerisms and engagement.

We also plan to introduce corporate-made memes which will appear every 5th post as advertising spots for potential advertisers, and we will (coincidentally) roll out a way to allow users to remove these ads via a premium subscription-based tier, a membership to the ChuckL Club, for $499 a month users who join the elite ChuckL Club will get an exclusive check-mark on their profile, which other users can see, and their comments will be automatically filtered to the top, regardless of how funny or even relevant they are.

We will also implement ChuckLChat ™, a humorous and lively chat-bot that will fill the void left by your platonic friends, who have long since uploaded their consciousness to the Live©.

Eventually, like all social medias, we will continue to bloat our application with so many features that the original charm and intent of the platform is meaningless, as it slowly loses its original charm in favor of converging into the jack-of-all trades amalgamation that is every modern social media platform.