The hacking concludes for this week so I’m happy to unveil the big project!

But there’s plenty more to work on after that.

And of course, there’s some YouTube too.

Topics for this week:

• Hackathon Submitted

• Hype Bid

• AWS Lightsail

• Reasoning in Action

• S Tiers for Business

Hackathon Submitted

We submitted our project for the Twitch Hackathon on Monday with 6 minutes to spare!

In true last minute, hackathon fashion, we submitted a couple code repositories, project details, a website and even a 3 minute video that gives an overview of the project with some slides for our long-term plans.

Not bad!

The submissions were due November 4 and the winners should be announced November 25 so I’ll let you know when we win!

If you want to see the project details, or maybe even give it a like 😉, you can check out our submission on Devpost:

Hype Bid

You might have noticed the name of our project is Hype Bid!

The original idea was to “invest” in streamers, and trade stocks based on hype their streaming.

Building something that tracks and calculates data for the hundreds of thousands of streamers on Twitch was a challenge that we initially couldn’t decide how to tackle.

But the big breakthrough was taking the fantasy football approach instead, narrowing the focus way down to a small group of users that could customize the experience to only focus on a few streamers.

A group of friends could sign up, create a league, pick the streamers they want to trade based on the our app tracks, and compete to build a portfolio of well performing streamer stocks.

I’ve built nearly all of that functionality from a code perspective, but the UI to interact with is a work in progress, and it will need a ton of testing once it has an interface to work with.

I did some work on that progress though!

We’ll need an app you can download on your phone eventually, which the team started building, but I’ve been building a quick web interface to work with as a prototype right now.

And it’s finally live!

Hypebid.tv

Connected to a real domain and everything.

But it is far from a real app.

Logging in sort of works, you can authorize your account with Twitch, but it doesn’t manage the session yet.

The Dashboard and Test tabs don’t actually do anything, but the Channels tab pulls in real Twitch channels and each channel has a graph page for their actual data!

Iterating on feedback will be vital for turning this into something real so the goal will be to get a basic flow ready for users to try out.

Progress!

There’s plenty of obvious things that are wrong with it but if you do check it out, let me know if you have any feedback.

Lots of work to do still so I’ll happily add new items to the to-do list.

AWS Lightsail

Amazon Web Services is responsible for some 30% of all of the internet so it’s hard not to host your web app on AWS at this point.

Lightsail is a specific AWS product that’s intended to make hosting on AWS simpler than managing all your connections to Docker images, databases, deployments, custom domains, pricing and scaling.

I think it would have been a quick and easy start to hosting our app if I wasn’t still working on expanding my Docker knowledge.

I didn’t realize I was building a Linux Alpine image and building it for the MacBook I’m using which the Linux instance I setup in Lightsail didn’t like.

Once I started using the proper build command docker buildx build -t tag:version -platform linux/amd64 deploying to Lightsail became much easier.

Past that small hiccup that blocked me for a couple days… my experience with Lightsail has been positive.

I’ve been able to make multiple deployments, setup a database to connect to, and connect the custom hypebid.tv domain.

I’d say, now that I know how to properly use Lightsail, I will certainly consider it for the next app I need to get hosted quickly.

Reasoning in Action

I finally had a clear example of the “reasoning” aspect which OpenAI has been promoting with their o1 models.

I’ve been a big proponent of coding with Sonnet 3.5 the last few weeks.

It’s helped build a good chunk of the Hype Bid app as well as a number of my other side projects, successfully giving me most of the code that I need to accomplish the task at hand.

But earlier this week I was going in a circle with Sonnet (post their new released version with computer control which I still need to try out).

I wanted to serve up the content of different pages in Hype Bid using HTMX and it gave me the correct answer of using HTMX to handle the content management aspect, but it had a tough time explaining how to do that.

It didn’t help that I was leading it down the wrong path because I was still thinking in terms of using Go to manage the content instead of HTMX, which was giving me issues.

After talking to Sonnet in a circle of wrong solutions, I decided to switch to o1-mini instead.

It took 8 seconds to think about the answer first, and then presented the correct implementation of HTMX logic, which once I saw it, I knew it was the right answer, making realize I was leading Sonnet down the wrong path.

But it was interesting to see o1 reason the answer first and explain why and then show how to implement the code as opposed to Sonnet just trying to do what I told it to do, even though it was wrong.

It was a great example of the “senior engineer” idea that Sam Altman has talked about in the past, expecting the AI to think and offer better solutions than blindly follow commands.

With that I’m quickly becoming best friends with o1-mini over Claude as my coding buddy but we’ll see how long that lasts before something new comes along!

S Tiers for Business

I’ve watched/listened to 2 different “social media for business” tier list videos now and YouTube seems to be the clear winner.

Alex Hormozi listed YouTube and Instagram as his top two “S Tier” choices for building a business and finding customers.

Devin Nash reacted to that exact video and had a lot to say in terms of Alex’s data and advice since Alex is coming from the perspective of finding clients that are making millions of dollars.

Devin works with much smaller creators and businesses so he offered his own tier list from that perspective, as a marketing agent working on a smaller scale or working with businesses that are just starting out.

But at the top of Devin’s tier list was also YouTube, and TikTok instead of Instagram.

Devin is more concerned with discoverability since you won’t have an audience starting out and YouTube and TikTok are great platforms for getting discovered.

Both platforms have their algorithms fine tuned to find the right person to watch the right video and stay on the platform as long as possible.

The general advice from two very different marketing experts is that your business should be on YouTube.

YouTube is S Tier for business.

And Instagram and TikTok may be as well.

I will certainly consider YouTube, and maybe TikTok and Instgram, as a strategies to grow any business I continue to work on, like say a certain fantasy market app for Twitch...

And that’s it for this week! Hackathon submissions, streamer stocks, AI reasoning and marketing a new business. All set for the future.

Those are the links that stuck with me throughout the week and a glimpse into what I personally worked on.

If you want to start a newsletter like this on beehiiv and support me in the process, here’s my referral link: https://www.beehiiv.com/?via=jay-peters. Otherwise, let me know what you think at @jaypetersdotdev or email [email protected], I’d love to hear your feedback. Thanks for reading!