Saturday, March 27, 2021

How deep tech skills can become irrelevant in the working world

In a chat with Gen X IT uncles, an ex-colleague had a grand theory. His hypothesis was that all the director-level IT guys he knows relied upon the coat-tails of another senior manager to climb up the career ladder. In fact, he asserts that skills don't really matter. So it's not what you know but who you know.

As we had the same colleagues in two US MNCs, I thought his observations are not too bad, but they can't be applied to all cases. The folks who are now directors in my generation cannot possibly be incompetents, they had some basic level of skill to supplement their political savvy.

From the last post, I sensed that younger tech workers largely disagreed with my analysis but the older guys agreed with my enlightened position on dealing with CECA and career obsolescence. 

But I need to make a clarification.

In this current economy, by all means, develop your technical skills. Computer Science graduates command the highest pay and employment rate in employment surveys and I do envy the jobs, my money-making buddies, Evan Koh of Stocks Cafe and Ivan Fok of Pyinvesting have. They are not just financially rewarding, but they involve a lot of data science and software development and involve negligible office politics. I was never to do work like this until I became self-employed and could build up my own web app development for my students. In the good old days, SGX staff would call me into a room to bash me for all sorts of horrible things my engineers did, and they did horrible things - like trip on a UNIX server power cable and halt trading in financial markets for half a day!

In my last meeting with Evan Koh, he said something really insightful about AI and software development. He actually felt that getting into the deeper end of AI ( tensors, back-propagation, neural network activation styles which I still struggle with ) may not be necessary for a future career because neural networks may be "pre-tuned" by someone else ( I suspect another AI ), many professional jobs should focus on practical application instead.

So here's my position :

  • As it stands, the economy does reward workers for having deep knowledge of tech. This is undeniable and developing it will give you a very high income without long and painful working hours. If you can do it, just do it.
  • However, tech workers should not think that it is your intelligence and skills that will carry you moving forward. Your rising star is similar to lawyers in the 90s - 00s. There was a supply-demand mismatch. If you look at the hordes of law students returning from England and Australia, even having a Guild to protect their interests now would still mean slowly declining salaries over time. 
  • You have to do a thought experiment - what happens if thousands of engineers descend into Singapore from India, Eastern Europe and China who can do your work for $2,000 a month. Are you still worth $8k-$14k? 
  • The tech dividend was paid out because of a political process that gave Aljunied and Seng Kang to the Worker's Party in 2011 and 2020. This constrains the ruling government from opening a CECA floodgate and benefitting the employer instead. The government's utilitarian calculus will always be that cheap engineers make Singapore competitive. Engineers must lose for everyone else to win. They'll do it until the MRTs break-down. Gen Y does not understand this, but Gen X does. Your money comes from disgruntled voters. 
  • You don't have to wait for CECA II to be signed to panic. You can chew gum and walk at the same time. While building your top-flight tech skills will take time, also build a financial portfolio with a focus on what happens when more foreign workers do arrive, even if it just to train us to be better engineers. Some argue that Silicon Vally landlords made more money than software developers during the boom. 
  • Even if you decide to do more tech, I see some developments that may shift the paradigm so abruptly, you can become obsolete overnight. In this regard, it may be safe to read up on Quantum Computing and understand its implications today. Evan explained this elegantly to me - an Algorithm that runs slowly with O(2^n) may become blazingly fast at O(n).
Before I end, I like to categorically say that I am not against CECA. 

In fact, I am for these floodgates to open via some new CECA, but my reason is that I have devoted the greater part of my life building a portfolio to benefit from that when my colleagues are looking for a pillar of support or patron to become directors in tech firms. 

I don't expect most Singaporeans to agree with me on this. I acknowledge that more competition and innovation is good for the overall economy and investments. But career dislocation in middle age is painful - I had a European IT colleague commit suicide after his contract was not renewed. 

[ Just for fun here's the list of IT certifications I held in my previous life: MCSE, MCSE-W2k, MCSD .NET, SCJP, LPIC-1, CCNA, CCDA, CISA, CGEIT, CISSP, PMP, Novell CNA, TOGAF, HP-UX Admin, AS400 admin ]


  1. Whoa, old-timer alert ... AS400, HP-UX, Novell ... hahaha my time too. You missed out on Cobol & IBM CICS. 🤣

    Solid hard skills are like a premium pedigree resume cum polished interview process. That gets you in the door & a fat starting salary.

    But whether you can grow the fat salary & perch on the top rungs of the totem pole depends a LOT on soft skills. The importance of hard skills degrade as you focus more on the big picture, manipulating and motivating fellow humans, & strategise on the outcomes that really matter -- profits.

    Those who can't, commonly term this as 3-legging, or balls carrying, or c*ck sucking.

    Not everyone is good at this. Especially East Asians who tend to be more deferring & don't like to talk so much & just put their heads down to get the job done & then do their own thing afterwards.

    That simply means they need to leverage on their strengths early on to build out their safety nets & passive income machines for the inevitable, whether that be in 10 yrs or 30 yrs. If you're in high-pay hard-skills, that's usually closer to 10-15 yrs unless you jump out of the high-pay hard-skills track.

  2. Not being able to do COBOL is one of my biggest regrets because when COBOL experts were retiring, I could have scored 5 digit contracts in my late 20s.

    But no... I was stupid enough to think that management was the next big step for me.