I hated the last part of 2024

A chaotic end to 2024.

On the first week I stopped posting, I posted on the community tab on YouTube saying I'll be back to regular weekly videos in December last year. But that never happened.

So here's all that happened, and to give you a couple of updates.

If you prefer to watch, this issue is available on YouTube as well.

If you've noticed, the studio setup looks a bit different from last year. Previously, I was using my phone's camera to do the face cam but from now on I will transition to using my Pocket 3 that I use for vlogs for my studio face cam setup as well. I hope this is a welcoming change, the new format should be better and nicer to consume, especially because now I'm directly facing the audience when I'm talking.

Endless Backlog

I am currently a Year 2 computer science student in Singapore, not sure if I had mentioned this anywhere before, but yes computer science, the then hype; now oversaturated course that nobody can seem to find a job at because of all the major layoffs in tech companies.

So school was tough last semester, and there was so much crazy work to do every single day. It especially peaked towards the end of the semester, which was around October to November. That was when I stopped posting. This was a chaotic domino effect. There was so much work from school to do, that it overwhelmed literally everything else. There was no time to do anything else except school projects.

It was crazy. The good news is I'm back, hopefully for a while.

And the domino effect? I have a whole backlog of things to do, which had me catching up all those other things after I was done with the school work. And some of the backlog included uploading the remaining shorts to YouTube, Instagram and TikTok. Space was a big problem. I couldn't record any more videos. I had 2 TB of video footage waiting for me to be cut into shorts on my computer. I didn't have time to look at it and I couldn't keep any more videos. I can't delete the footage until I've cut the shorts because I needed the raw footages to still be there in order for it to work.

Burnout

To put it shortly, I burnt out. Putting out regular weekly content isn't easy, especially when I am juggling between so many different things.

This never intended to be a travel vlog channel in the first place, it just so happened we were travelling a lot nearing the end of last year and I decided to start vlogging and recording down the things we did while we were travelling. Eventually, we had so much vlog content that this technically became a vlog channel.

It takes like half a day minimum to edit an entire vlog in one sitting. Despite the assist with AI integrated features to help making editing easier, it still takes up a lot of time and effort to edit a video. Considering my videos don't really have any special editing, intros or outros already.

I had so much other content planned but none of them eventually got to see the light of the tunnel because I was so busy with pumping out the weekly travel vlog content.

You will be seeing a lot more videos in this style as well, because this was the original intent and purpose that I created this channel for.

And I'm just doing way too many things in parallel, that I don't really take breaks. While everyone else is enjoying their trimester break, I'm back at part time job doing work. The only breaks I really took was only when we travel. But they aren't really considered breaks either, travel is a really tiring thing, even more tiring than getting off a long day of school or work. So much planning and doing involved.

The last batch of travel vlogs I have is the Korea vlogs. We went in October last year. As soon as we came back, everything was chaos. Not sure when these vlogs will be up, there's still a heck ton of editing yet to be done. It will probably be a series of 5-7 videos. Wanted to pump them up for vlogmas last year but I just didn't had the feel to do so much editing work.

And vlogging in itself is a very tiring thing to do. Because when you are already travelling, there is so much planning and things to look out for. Add in the vlogging, it's twice as tiring. Honestly, a huge respect to those who vlog consistently on a regular basis. It's really not easy. But maybe if you do it regularly, you get used to it and it becomes easier. It's not really my type of content, unless I'm travelling. It's so tiring to be talking to a camera non-stop the whole day when travelling is already a big burn in energy.

Well, we've been travelling a lot and I will be making a trip the UK this Summer for a few weeks for a school exchange. Hopefully I will be able to pump out some nice vlog content while I'm there.

Decosmic & Been

For Been and Decosmic, the 2 products I have been over the last year — I will give an update on these 2 separately in another video. It's a bit too long to cover it here.

2025

I learnt a lot the last year. In early February, I started building Decosmic. It was a lot of lessons learnt and problem solving. It's not easy to build a startup as I thought.

I had a microstartup built to 470k users when I was 15, but that was in a gaming community. The reach was perhaps easier, and I realised that launching in the real world is tough. Building the product itself is the easiest part, there's so much more to it, the marketing, the legal, the admin. And there's so much competition and everything keeps revolving, one step late and everything is over.

And especially in the AI space, it's so oversaturated. It's hard to find a niche, something someone hasn't already done before. Everything you can think off, somebody out there has already milked it. And most of the problems with building AI products, is finding a use case for it. In 2025, we are still figuring out the proper use case for AI. So far it has always been still a chat box, others pulled it out of it and gave them access to tools, context, memory, and these became what we know as agents today. But it's still not good enough to be fully autonomous at this stage. And everything needs so much customization, it's all a workflow editor; drag and drop, change the input and output.

AGI is still far away from 2025. We're now doing a lot of inference time scaling, which in other words making the model think before it responds. We are seeing a lot of thinking models being released after o1 kickstarted the race. A few days ago, DeepSeek R1 got released, a fully open source thinking model and from what I heard from the community testing it, the performance is on par and sometimes exceed o1's performance.

A New Chapter

But the good news is that I didn't just disappear and did nothing for the last 2 months. I was laying a lot of ground work to prepare for this year. Because last year was just a lot on experimentation and figuring things out. But I think I've got the hang of it, I roughly know the direction I want to head towards and this year will just be focused on that.

So yesterday, I launched Update Night, a newsletter where I bring you the under covered tech stories nobody else is talking about because they are buried under the headlines. Doing the first issue of the series made me realise how much news there were in the community that most people don't even know about, because the big news outlets will never cover them unless they are from a big company.

But it also made me wonder how long would it take for me to burn out. Doing this every week might be doable, but when there is just so much developments every week especially in the AI space alone, gives me so much work to write out one article. And I don't want to miss out anything. Perhaps the next issue will be condensed and tuned to the more important news.

And I've recently incorporated a business in Singapore, it's called Base 7. This was a long time coming, it was my dream. And I still can't believe it. But incorporating is just the beginning, getting to profitability and product market fit is the ultimate destination.

This company builds in-house products, basically I'm putting everything I'm building under it. And we're looking into expanding into an agency as well, to build full enterprise software solutions for clients. Feel free to drop us an email if you have anything in mind, we will give you a free quote.

So we have about 3-5 new in-house products in development or rather some of it is still under planning. If you include the ones we've already announced.. but not yet officially launched or completed (which is Decosmic and Been), it's about 7. I can't wait to launch them as soon as they are ready.

Restart

Unfortunately my channel has stood still for a few months, the exploding subscriber growth stopped at the last video. We still haven't cross 1,000 subscribers yet. We're so close but yet we're so far away. Surely, we will cross this line this year.

I just love creating things. It's my superpower, as X wrapped says.

It can be seen as a bad thing to have nothing finished but having an endless flow of concrete ideas I can actually provide value with is amazing. I love having things worked in parallel. Even though it can be seen as you can complete 1 by focusing all your time on that one product instead of doing so many, working on multiple projects at the same time gives you a different angle to every piece of work you create.

All these combined together you learn twice, triple times as fast as you would than what you would have learnt doing one. And doing all these in parallel, means you can immediately apply all the lessons and mistakes you learn across everything you're building, reducing technical debt.

But I'll cover more on this in a future video.

I've become accustomed to creating opportunities for myself. Instead of waiting for them to come, I create them. And a lot of the things I will be doing will all revolve around this idea.

So a lot of exciting announcements coming the next few months and later this year, so keep an eye lookout for that.

I'm excited.

Cheers,
Jay