Stopping my YouTube channel until further notice

 

For the next few weeks, I've decided I can't sustain my YouTube channel until I satisfy my craving for AI to assist me with my business initiatives. 

Across LinkedIn, people are fairly confused about how to pick up enough AI skills to remain relevant.

This is also a problem I'm struggling with, but it's really worth the struggle. Over the past 48 hours, AI also converted all my course materials into a textbook that may be published in the future.

There are simply too many questions on my mind.

a) What would happen if my industry woke up and started using AI aggressively? 

Will investment courses become more AI-driven or human in character? I think Dr Wealth's position is that investment training will become more narrative-driven and human, but I'm inclined to bet otherwise. But we can advance both agendas in one go, make the course more powerful and intuitive at the same time, while preserving the student's human judgment. I don't really value things like emotions and empathy because they don't make me much money in the markets, but I'm open-minded and willing to change.

b) How will content creation change?

Ideally, I should be so productive that I can work on my gigs and create more content on my channel, but I need to understand how professional content creators integrate the latest AI tools into their channels and whether there's a way to increase revenue at the same time (without generating AI slop). This means I need to watch a lot more YouTube videos to level up my skills in this area. 

c) How do we even start an aggressive campaign to up our AI skills?

Anyone who can crack this can make a lot of money. At the moment, I'm just clinging to Claude Anthropic's courses and trying to get as many certificates under my belt, but what I really crave is to develop several consequential personal projects to learn about agents and level up my course materials. I want to reshape my CV into something a 25-year-old might have, a portfolio of tangible projects and proven results. 

The culmination of all these changes means that I can, at best, focus on my business and gigs in the next quarter while I invest most of my time learning about AI to see how far I can push it. 

And since I barely do any conventional work, there are very few efficiency gains from AI unless I dream up and create new products like simple mobile apps, a personal productivity tool, and at least one D&D module for 4th-level PCs. I'm cognisant that in such a situation, FIRE can lead to complacency as the flow of dividends just provides so much incentive to let the world just pass by. At this point in time, I'm glad that I still have a few gigs to run and would actually be teaching IT to poly students next semester.

If you are a reader and wish to help me out, send me details about when Gen Z are meeting to discuss things like Open Claw or Claude. 

I'd like to join to learn, by osmosis, what these young engineers or computer scientists are thinking.

Using Claude Code to upgrade my business

Last week was busy, even though it was the school holidays, with one SIAS course followed up with a preview. At the same time, I've managed to sign up for classes for Athropic's Claude, which covers a mix of high-level philosophical lessons on incorporating AI into human work and just one technical introduction to Claude Code.

And what such a small difference would have made!

Within 30 minutes of barely understanding how Claude Code worked, I began using it to improve the Early Retirement Masterclass web app.

For reference, here is a screenshot of my old web app, which is still in production for my next course run.

After using about $1 worth of tokens, this is the latest version of my website.

And I've only just begun in website design, and UX is my greatest weakness. It's an area I'm not really interested in learning, and now I have a person who can adjust font sizes and change colours for about $0.30.

But it goes deeper: AI can collaborate with me to improve my website. I initially wanted to incorporate LLMs into my tool to pre-fill the forms my students use. It was the AI that suggested making direct API calls to Yahoo Finance instead.

So no more broken English when students mangle the main business of the stocks they are investigating in our final portfolio reports.

At this stage, every investing trainer needs to ask themselves the question:

What kind of investigation into a local stock can be enhanced by AI, and what kind of judgment call had best be done by human beings? Claude places this under Delegation in their 4Ds AI framework. 

My program has a fully fleshed-out Markdown file to automate stock valuation. Here's one of many paragraphs produced for YZJ Financial Holdings.

I've designed the analysis to rely solely on downloaded financial statements and avoid referencing human analyst reports until the last step. I find that such reports are superior because they are more willing to make sell calls and price a stock as brutally and realistically as possible.

The old web app will still be used in next week's Batch 41 program. The new web app will be deployed to the cloud after the course. 

Finally, putting such powerful tools to work for any business or solopreneur has implications for the economy. I prefer to wait for these social scientists to think through the implications of this. There are just too many applications I'm dreaming of to do any navel-gazing on jobs, governance, and policy implications.

I only know that dividends can be converted into AI tokens, which can lead to larger businesses and more passive income, enabling better AI tools and more tokens, provided you are ready to retool yourself as a dreamer.

The time has come to be a maker and not a taker.

Students can refer to Dr Wealth for a refresher.

Anyway, if you are wondering why I've not been creating new content. 

The above is the reason why.

Would you bet on CS salaries over the next 10 years ?

 

I recently turned down a bet, but I felt it was an interesting one worth sharing here.

In an age of AI disruption and the over-hiring of software engineers over the past few years, my friend offered me a bet that, after 10 years, the salaries of Computer Science graduates would not keep pace with Singapore's inflation rate.

That's a very nice bet that if we have some kind of Polymarket in Singapore, I think readers of this blog can give it a try to see whether they are right.

Initially, I felt that a 10-year horizon was too long, and I actually offered the same bet with a 3-year horizon, but my friends refused to pick it up. So we ended in a stalemate.

So here's the funny thing. If we kept the 3-year horizon, my friend has a good chance of winning the bet because the NUS CS grad salary actually fell from 2024 to 2025. Tech overhiring is real. But if we tried a 10-year horizon, my odds would increase tremendously, simply because wage growth here exceeds the inflation rate. ( In small part due to the MAS approach to manage the SGD )

And I can go on. Some economists believe that the demand for software engineering would actually shoot up over the next few years because of this theory known as Jevons Paradox. With IT agents and cheaper tokens, many people will come to believe they can spin off companies and develop new SaaS solutions using an army of coding agents. One possibility is that a huge industry of software engineers is required to pick up the mess. Another is the rise of an agent maestro, a totally new career that requires domain expertise and software engineering skills. I strongly believe in this thesis because even Sheng Siong is looking to hire SWEs.

So after some examination, I think this is a good bet for me, provided we can narrow this down to NUS salaries. If we expand this to the median CS salary, there will be a serious problem because that represents a CS degree over 10 years. If the Arts faculty launches a Bachelor of Arts in Artificial Intelligence (which I consider an inevitable move to keep humanities relevant), do I need to include the income of an Arts major?

But at the end of the day, 10 years is too long. 

A good bet does not need to be financially significant, but it should exact a psychological cost or an actual ego boost.

I suspect the bet was set up because there was some animus against CS grads on my friend's part, and I can't provide an ego boost by losing because I'm basically a freelance hobo, but I also cannot gain an edge if I win. My only mischief is that I actually like CS grads because they debugged my trading algorithms and upped my Sharpe ratio from 1.2 to 2.

I think a better bet would be to consider the future of CS grads against some kind of Compliance career, which I think lies at the root of this animus. But this is not something I care to take up because I actually benefited a lot transitioning from IT support to IT governance. Harsh truth: Society does reward the box-tickers in the corporate world. 

I had more pay and less work.

The worthlessness of box-tickers in the corporate world is amply discussed in cultural anthropologist David Graeber's Bullshit Jobs, and believe me, there's no way their remuneration is going down in the age of AI; they will start gaining expertise in AI governance and find some way to become consultants in some ISO-676767 AI certification - if ever to keep being a pain the ass for hapless engineer.

But the rage could be philosophical. 

In Ayn Rand's philosophy, I suspect she loves CS grads as much as I do because they seem closer to the hero in her books, like John Galt or Hank Rearden. But in her novels, compliance professionals are the hyenas and parasites; what have they produced other than getting in the way of the people who do the real work? 

While I'm not betting on this, I have secured teaching stints in both the Law and IT departments at a Polytechnic, and I'm a parent, too. So I might want to see how starting salaries for lawyers and CS grads will evolve over the next few years.

So expect some tracking on this blog moving forward.

Happy Chap Goh Meh !

 

I had to miss out on a Valentine's Day post because of the intensity of my marking workload and grappling with changes in how investment courses are conducted in the age of Agentic AI. But since today is Chap Goh Meh, which is the Chinese equivalent of Valentine's Day, I shall dispense with dating wisdom on this blog today instead.

Clearly, the context in the dating space is grim, as Singapore faces an all-time-low TFR of 0.87. I'm not sure whether it's coincidental that markets are at an all-time high. But the dating market is already quite broken.

a) Maybe it's time to ditch dating apps

As I consume a lot of social media poison on Instagram, I sometimes get hilarious video snippets that explain what happened on their dates. And Gen Z guys on dating platforms these days are losers. Insisting on splitting the bill is fine, but more importantly, not being able to articulate career goals and bragging about commercial pursuits that mark you as a loser, drop-shipping or crypto trades.

I think there are many good guys out there, but maybe a dating app isn't the best place to find a guy. I still believe that to find guys with a high IQ and high conscientiousness, you need to search in institutions of higher education, then let assortative mating take care of itself. 

b) Singlehood is the biggest hurdle a guy needs to overcome

Middle-aged wives can be brutal. When I was talking about online dating and younger Gen Z norms, they brutally reminded me that men these days no longer compete with each other; they compete against singlehood - and singlehood is winning.

So basically, a lot of women realise that singlehood is better than getting attached to a guy with the same attractiveness score, so they stay single and have a good life dodging kids and domestic stress. 

I'm not sure that this argument applies just to women. While I know a lot of guys who can't apply themselves and have no sensible way to navigate their lives, I think single men have plenty of enjoyable options, too, with AI tools and entertainment. Fembots are a few years away, if not here already.

It was Boomer male who invented the phrase "Why buy the whole cow when all you want is just a glass of milk?"

c) There are a lot of beautiful women, but only a few successful men.

If I can advise Gen Z men, this is something I would say. Actually, if you are logical about it, this is a completely objective statement that is not misogynistic at all. 

First of all, we have to accept that women are pickier, so their subjective standards of success will always lead to a smaller population of men, whereas men's flexible standards of female beauty can be quite wide, so the population of beautiful women will always be men who meet women's standards of success.

What does it mean? Everything that I've been talking about. Guys need to work on themselves and climb the ladder of success. 

Many beautiful women are waiting for them once they reach the top.

Just ask the PropertyLim people.

Happy Chap Goh Meh!

Why no new content since the new year?

 

Ok, I've been unexpectedly busy for the Chinese New Year. But I'm cautiously optimistic that things will pay off over the next few months.

So, the biggest item on my plate is the upcoming talk I'll be giving for SIAS on small and mid-cap companies. In hindsight, creating slides for this one-hour lecture is harder than I thought because I have to distil all my ideas on small-cap investing into a set of slides. And this goes beyond what I ordinarily do for my investment courses. 

I'm not even sure whether doing this might be the beginning of a new course with Dr Wealth, but there's plenty of new material to incorporate into my greater course slides.

You can access the link here to register for it: https://portal.sias.org.sg/Event/EventDetails?Id=f4c2b063-390d-f111-8406-000d3aa337f8&EventName=Unlocking_Opportunities_in_Small-Mid_Cap_SGX_Stocks

The second item on my plate is that, thanks to a blog reader, I have been introduced to the IT faculty at the Polytechnic where I teach, which might see me conduct some technical courses in addition to the law courses I currently teach in the evenings. For one thing, I'm not going to turn away CPF money, which I still want to accumulate after 13 years out of conventional employment, but the idea of someone teaching hardcore analytics and data science courses alongside law modules is not something I want to miss out on in my resume. 

The price of being selected is also high, as I have to teach myself at least two software packages that I'm unfamiliar with. It seems that my days of avoiding Power BI in favour of hardcore coding in Python might be finally over. 

And of course, I have to prepare for Batch 41 of the Early Retirement Masterclass, in end-March, which will launch a Markdown skills file for students to analyse local stocks. 

As I'm still working on slides and prepping for workshops, I haven't had time to digest geopolitics, so I don't have any insight into what will happen in the markets tomorrow. But I do know that my All-Weather Portfolio Masterclass students should be happy since we've been into defence stocks and commodities for the past 6 months or so. 

Maybe if I can complete my slides for SIAS, I might go that new Spa in Perrenial Business City....