Ministry of Manpower has just released its annual summary of the trends in job vacancies in Singapore, together with a list of the professions seeking the most positions that companies seek to fill.

The survey is very comprehensive, having covered private sector companies which employ at least 25 people and the public sector comprising government ministries and statutory boards.

A total of 16,800 establishments employing 2,168,900 employees responded to the survey, with a response rate of 89.0%, between 15 September 2023 and 19 December 2023.

Here’s the breakdown of the findings:

New job creation picks up

2023 was the best year for new job vacancies since 2018 when the figure was first reported. Nearly half, 47.3 per cent, were new positions.

The slight dip in 2022 is somewhat understandable, given the post-pandemic hiring spree, which was all about filling the jobs that were lost in the pandemic, but the bump vs. pre-pandemic levels shows high levels of optimism even as Singapore economy is now employing more people than ever.

Where are those new jobs created?

Unsurprisingly, the industry with the highest proportion of new openings remains the ICT sector, which is forever hungry for more IT talent. Nearly three out of four openings were created over the past year. It’s no surprise then that even fresh graduates can expect starting salaries of $5000 to over $6000 per month.

In second place comes the still booming construction sector, although, with cooling measures taking greater effect, this may taper off in the coming year or two.

Interestingly, food and beverage services and retail trade are also hiring more than they are simply replacing, followed closely by transportation and finance, the bedrock of the Singapore economy.

On the opposite end, real estate is showing signs of saturation, as is accommodation and arts & recreation, still seeking to fill the post-pandemic gaps, as travel to Singapore is back to 2019 levels and set to grow in the coming years.

PMETs in demand

It’s quite remarkable that as recently as 2017 non-PMET vacancies still outnumbered the PMET jobs on offer. Within a decade since 2014, the proportions have reversed, showing a progressing professionalisation and uplift of Singapore’s labour force, as more and more jobs require more advanced expertise (driven mostly by the tech sector).

Obstacles to employment

We often hear that many seemingly good, decent jobs for professionals are picked up by foreigners, but there are certain reasons why local residents (citizens and permanent residents) either refuse or simply can’t fill them.

And, as it is often cited by employers, the main reasons are the lack of skills or experience, with insufficient pay being the cause in just 1/3rd of all cases of vacancies remaining on the market for more than six months.

Degrees matter less and less

Reflecting global trends, another revolution in Singapore’s labour market is the rapid decline in the importance of academic qualifications.

Of course, this was never a major factor for most non-PMET jobs, but since 2017 it has very rapidly become a non-issue for PMET positions as well.

From just nearly 60 per cent seven years ago to just 33 per cent today — this is how rapidly the importance of your formal education has declined.

For two-thirds of all vacancies, even among professionals, employers will consider candidates who do not command the necessary basic education for the role, provided that they can offer other valuable traits. Businesses place greater emphasis on experience and real-life skills rather than paper qualifications, and that seems to be both fairer candidates and better for the economy.

Top vacancies with expected salaries

According to MOM, educational background is not the main consideration chiefly among software and media developers, management executives, business development managers, sales executives, and even some teaching and training professionals.

Incidentally, all of them are on the list below:

Top 10 PMET Vacancies

Rank	Occupation	Range of wages offered
1	Software, Web & Multimedia Developer	$5,000 to $8,505
2	Teaching & Training Professional	$3,080 to $8,580
3	Commercial & Marketing Sales Executive	3,000 to $4,200
4	Management Executive	$2,400 to $6,250
5	Budgeting & Financial Accounting Manager	$7,000 to $10,000
6	Industrial & Production Engineer	$4,000 to $6,500
7	Management & Business Consultant	$5,000 to $8,000
8	Systems Analyst	$5,500 to $9,663
9	Registered Nurse & Other Nursing Professional	$2,730 to $5,000
10	Business Development Manager	$6,000 to $10,000

This proves that not only does the labour market reward skills and experience more than paper qualifications, but that education in Singapore should place greater emphasis on developing practical competences from a young age to future-proof the new generations for growing international competition.

After all, if you don’t evolve, you’re not only standing still—you’re falling behind.

Following many months of controversy surrounding the thus far rather secretive Elon Musk venture, Neuralink, the company has run a 9-minute live broadcast on X with the first patient who received one of its implants — and the reveal was more than promising.

https://t.co/OMIeGGjYtG

— Neuralink (@neuralink) March 20, 2024

29-year-old Noland Arbaugh suffered a broken spine in a diving accident eight years ago, leaving him quadriplegic: completely paralysed from the neck down. The consequences and resultant dependence on others for just about everything need not be explained here, of course, so every invention that permits the patient to communicate with a machine is a life-changer.

Neuralink’s stated goal is “creating a general-purpose, high-bandwidth interface to the brain”, which would allow all humans to communicate with computers directly via thoughts rather than typing or talking.

That said, the obvious candidates for early testing are those most in need of such a technology to return to doing most basic things in life.

In the long run, it may even be possible to use this innovation to fully restore motor function in paralysed patients one day, as per Elon Musk’s comment under the video:

Long-term, it is possible to shunt the signals from the brain motor cortex past the damaged part of the spine to enable people to walk again and use their arms normally

— Elon Musk (@elonmusk) March 21, 2024

As of now, however, even doing as little as being able to control a laptop is an enormous improvement in the quality of everyday life.

Using the Force

So, how does it work? What does it feel like? Noland explained that it’s like using the Force in Star Wars movies — you just command the cursor to move where you want it to wirelessly, with nothing else but your thoughts (and no cables sticking out of your head, like in the past).

The process is a bit more complex, of course, and initially involves visualising movement and imagining what you would do to make the cursor move to the place you want, while the implant learns to translate your intent into motion on the screen.

But once it’s been dialled in, it becomes intuitive. So intuitive, in fact, that once he got his implant under control, he stayed up for 8 hours straight playing Civilisation VI while lying in his bed.

He’s also playing online chess with other players on the internet and is learning Japanese and French.

Long sessions at the computer were previously difficult, since lack of motion for many hours causes pressure sores and requires assistance from family to move his body around even in the wheelchair.

Right now, however, he can stay comfortably in bed, relying on himself to control the computer with nothing but his thoughts. The only limitation is the implant’s battery, which needs to be recharged periodically.

He admits that not everything was smooth sailing and it’s still a work in progress, but the impact it has had on his life is profound:

“I would say that we have run into some issues. I don’t want people to think that it is the end of the journey, there’s still a lot of work to be done, but it has already changed my life.”

Noland Arbaugh

Even the initial surgery, while invasive, went well and Noland was out of hospital after just one day.

The same cannot, probably, be said of the monkeys Neuralink is using to develop another implant: one whose aim is restoring sight.

Monkeys see again?

Following the broadcast Elon went on to introduce another product, called Blindsight, dedicated to patients who had either lost their sight or were born blind.

I should mention that the Blindsight implant is already working in monkeys.

Resolution will be low at first, like early Nintendo graphics, but ultimately may exceed normal human vision.

(Also, no monkey has died or been seriously injured by a Neuralink device!)

— Elon Musk (@elonmusk) March 21, 2024

As he explained in a short video clip attached to the post, “the visual part of the cortex is still there”, so that even in extreme cases of blindness at birth, sight may eventually be restored.

The implant is currently being tested in monkeys and, per Musk, “is already working”. The billionaire took the opportunity to jab back at his critics, after allegations of animal abuse and numerous deaths of test subjects (monkeys) cast shadow on Neuralink’s practices.

That said, if the early results are anything to go by, then even if that happens occasionally, it is perhaps a small price to pay for the life-changing effects the company’s technology may have on humans.

Featured Image Credit: Neuralink

Disclaimer: Opinions expressed below belong solely to the author.

On November 30, 2022, the day OpenAI released ChatGPT to the public, we were thrust into a completely new reality — though not an unfamiliar one, given how many sci-fi novels and movies predicted that one day we would be conversing with intelligent machines.

Launched around the same time, diffusion models showed us that computers can learn to generate imagery at will, followed by increasingly more sophisticated video footage, as recently exhibited by OpenAI’s Sora.

Quite overnight, computers have been granted powers to do things that thus far only humans could do — create, speak, design, draw, and write.

Since then, the AI craze has taken over the world. We’re discussing how and when robots will take our jobs, making millions of people redundant, before conquering the planet and threatening our very existence.

On their way to total dominance they appear to have already conquered the world of business. A seemingly endless stream of billions of dollars is being poured globally into companies building competing AI models and the hardware necessary to run them.

And everybody wants to get a piece of the pie.

Not only are we now getting laptops or smartphones with AI features, but also AI vacuum cleaners and lawnmowers. Millions of inanimate, electronic devices seem to be acquiring reasoning skills which are expected to turn our world upside down within a few short years.

Some experts in the field make bold predictions that Artificial General Intelligence — machines with reasoning capacity equal to human beings — could arrive in as little as three to eight years.

Even Nvidia’s CEO, Jensen Huang, predicted that AI would be able to pass all human tests within five years.

However, if history teaches us anything, we might be in for a much lengthier wait.

Law of diminishing returns

With the dawn of AI-enabled washing machines and refrigerators, I think it’s worth to pause and think if we’re not getting ahead of ourselves.

The nature of technological development is that it typically starts with a breakthrough before gradually tapering off as evolution and refinement of the new invention inevitably take more and more time. There is no reason why AI should be different.

Think of it this way — the first slice of pizza usually provides the most satisfaction. You get quite a lot of it with your second or third slices as well. But at some point in satisfying your hunger each bite you take will provide less and less gratification — less “marginal return” or “marginal utility”.

Remember how 5G was promoted as the next big thing in technology? But how many of us have really noticed any difference? 4G was good enough for the vast majority of everyday uses. 3G was a huge leap in performance over 2G, which in turn provided the largest marginal leap by making the web reasonably accessible on mobile devices for the first time.

Life-changing technological developments have always needed enormous effort, even if the path was clear to engineers and investors.

They also require a lot of time.

It had taken 25 years for general-purpose computers to leave military compounds and universities, making their debut in people’s homes in the early 1970s, and another 40 years before reaching global saturation.

These days, they are overtaken by personal, mobile devices such as smartphones, which can perform many of the same functions.

On the topic of mobile phones, the world’s first launched in 1983, but it had taken another 20 years for it to become an ubiquitous device even in the developed world.

Today, most humans have a smartphone in their pockets, a technological “revolution” which has, however, needed four decades to fully unfold.

Or let’s take a more recent example: Tesla.

The trailblazing EV company launched the Autopilot driver-assistance system for its cars in 2013, and ever since then, we’ve been promised that self-driving vehicles were just around the corner.

In fact, Elon Musk has made so many failed predictions on the topic, that it has a dedicated Wikipedia page.

Many, perhaps even most, people bought into these visions because the original Autopilot was already very good. If we can teach the car to drive itself in most situations, how difficult can it be to improve it so that it can drive us reliably all the time?

The answer is: very.

In fact, the parallel between self-driving and AI is particularly relevant, since both technologies aim to teach machines to behave and think like a human being.

And it’s why AI is likely to struggle going forward just as self-driving technologies have.

It’s easy to teach a car to reliably change lanes, adjust speed or avoid obstacles in most conditions — but it’s very difficult to make it adapt to those which are less predictable and flexibly judge the changing situation on the road.

In the same way, it’s far easier to make a machine learn to mimic human speech and writing in most situations, giving it an impression of intelligence. However, it is far more difficult to actually make it think.

If you don’t trust me on this you don’t have to look far for confirmation of the slowdown in AI development — just look at the progress of the past 15 months.

How big was the leap between ChatGPT 3.5 and 4?

How many new, important features have been added in the past year? We’re 15 months into the AI revolution and yet nearly all that the “intelligent” chatbots do is the same as on day 1: summarise documents, provide simple suggestions, write or rewrite documents, answer questions, write simple code.

It’s the same among image generators, so universally vilified by artists and designers.

After initial improvements that made images they produce more usable and with the recent addition of better handling of text, their outputs haven’t exactly improved by leaps and bounds.

Most are still limited to a quite paltry resolution of 1 megapixel (1024x1024px) and have problems with accuracy or consistency.

Making that first impression of doing something out of nothing was far easier — and far more noteworthy — than spending months grinding just to make sure AI-generated humans have five fingers in each hand.

Former US president, Donald Trump, experienced it first “hand” two months ago when he posted an AI generated picture of himself praying with… six fingers on his hands.

Even after well over a year since the launch of AI image generators to the public, they are still struggling with things that are obvious and very basic to humans. Fixing them keeps consuming significant effort — quite unexpectedly, considering how we are being told that AI is about to outsmart us.

This is where AI revolution is currently at: its “grind phase”, where thousands of engineers around the world are working very hard refining things that should be fairly easy if the technology was really advanced.

Making sure the “intelligent” models don’t hallucinate, that they provide accurate outputs at all times, that they don’t mislead or insult their users. A million details that still require human sweat to get right.

Growing inputs produce increasingly smaller incremental outputs.

Yann LeCun, Meta’s lead AI researcher, highlighted the problem with the current approach upon OpenAI’s reveal of Sora, its text-to-video generating tool: it can’t really think or perceive objects, scenes, people, animals it creates.

Let me clear a *huge* misunderstanding here.
The generation of mostly realistic-looking videos from prompts *does not* indicate that a system understands the physical world.
Generation is very different from causal prediction from a world model.
The space of plausible videos is… https://t.co/gBCdTOnDbp

— Yann LeCun (@ylecun) February 17, 2024

It merely tries to accurately predict what the next pixels in the sequence should be, without understanding what it actually depicts (even if it’s remarkably convincing at it).

It’s a really compelling approximation rather than an intelligent creation — and still only a prototype that isn’t expected to launch anytime soon.

Slow is good

We’re at a point where producing smaller and smaller improvements is going to require increasingly larger investments. Such is the reality of progress in every domain of life, though (think of how much more difficult it is to set a new World Record in 100m sprint today than it was 50 years ago).

We may already be past the “wow!” moment that the launch of ChatGPT was. Every new AI service or product is likely to be less impressive, though still important and, ultimately, transformative to our lives (just like all technology before it).

The inevitable slowdown isn’t a bad sign. Quite the contrary, it’s how it has always been. The first iPhone was a bigger breakthrough than iPhone 15 but you wouldn’t trade the new one for the old one, would you?

Since we’re still in AI’s early days, we may occasionally see quite large leaps of progress, especially as the technology enters new fields.

But we also have to accept that even people who are driving the revolution say that it will require trillions of dollars and years of investments, making it one of the most expensive global endeavours ever undertaken by man.

So, we should neither be deluded by very human-like conversations with AI bots into thinking that smart androids are just around the corner, nor by the doom-and-gloom predictions of the naysayers, who claim that humanity is about to end.

We’re likely at least decades away from creating a truly thinking machine — but it’s unlikely to be the Terminator.

Featured Image: Nvidia

Disclaimer: Any opinions expressed below belong solely to the author. No information published below should be treated as financial advice.

Despite growing competition in the e-commerce segment, Sea Ltd. has managed to achieve its first profitable year since going public in 2017. A fairly modest US$162 million in net income last year compared favourably against the loss of US$1.7 billion in 2022.

The stock market reacted positively, jumping as much as 11 per cent during yesterday’s session, before closing the day at 5.6 per cent, as some investors cashed their gains out.

This bump has pushed the year-to-date rally for Sea Ltd. stock beyond 40 per cent over just 2 months, reflecting a return of optimism about its future.

Shopee fends off competitors

Interestingly, the positive result was achieved despite a reversal of cost-cutting policy, which saw sales & marketing expenses plummet in late 2022.

In the 4th quarter of 2023 the company spent more than twice on promotional activities than it did the year before, approaching US$1 billion, driven mainly by the response to competition from Chinese companies like Temu.

This has helped to boost quarterly orders by 46 per cent and the Gross Merchandise Value by over US$5 billion compared to the Q4 of 2022.

Thanks to strong quarterly performance Shopee managed avoid statistical stagnation for the entire 2023, as without the additional 800 million orders in Q4 it would have struggled to show annual growth.

This is likely to become a recurring theme for Sea for the foreseeable future, as it tries to balance the need to spend enough to drive purchases and staying profitable in a time of restricted access to cheap capital.

In this context, perhaps the best news to come out of the company is that it has managed to grow its cash reserves by another US$1.6 billion this year, showing that liquidity should not be a problem (although it still can’t spend nearly as much as it did before the onset of global inflation in 2022).

Garena shrinks in half…

All of the above is particularly good news given that Sea’s cash cow, the digital entertainment arm Garena, has suffered a slump in the post-pandemic world, where people are no longer stuck at home playing video games.

Revenue in the segment fell by over 46 per cent, from US$948 million to just US$510 million for the last quarter, and by 44 per cent, from US$3.87 to just US$2.1 billion for the entire year.

All of these figures are below 50 per cent of what the company pulled in from gaming in 2021, when annual revenue topped US$4.3 billion, at the height of COVID-19 lockdowns.

…but SeaMoney fills the gap

Sea Ltd. is in a constant state of flux between its three constituent businesses. When one struggles others pick up.

It all started with gaming, which evolved into e-commerce — that now provides the bulk of the revenue and is likely seen by investors as the most valuable part — and digital finance, growing at a decent pace, having provided US$1.8 billion in revenue, an increase of 44 per cent over 2022.

This means that even as the composition of Sea’s revenue flows changes, the total figure keeps climbing.

In 2022 Garena brought in three times as much money as SeaMoney did, but a mere year later they were almost neck and neck.

If the trend continues (as it might given the stagnation in mobile gaming) the business that Sea was founded on may be the smallest of all by the end of 2024.

Nevertheless, Sea is seen mainly as an ecommerce company and the fortunes of Shopee are likely to dictate how it fares in the next few years, before digital banking can establish itself as a potent money maker. By then, digital entertainment might become a side note in its books.

Disclaimer: Unless otherwise stated any opinions expressed below belong solely to the author.

Nvidia, the long-time graphics card company turned AI giant, joined the US$1 trillion valuation club last year and is now just US$50 million shy of US$2 trillion, following the surge in demand for its accelerator cards that large language models like ChatGPT are trained on.

Its multibillionaire founder and CEO, Jensen Huang, is now one of the richest people in the world, with a net worth of close to US$70 billion, and one that is most sought after to speak about the future that his company is helping to build.

He was one of the headline guests at the World Government Summit in Dubai, which concluded a few days ago. At the event, he spoke at length about the future of the world in this new AI reality, including the impact it’s going to have on our careers going forward.

According to Jensen, the mantra of learning to code or teaching your kids how to program or even pursue a career in computer science, which was so dominant over the past 10 to 15 years, has now been thrown out of the window.

You’re already a programmer

Perhaps a bit paradoxically, the recent achievements of the IT industry are leading it to automate itself first, thereby reducing the need for tech experts and the number of tech jobs in the long run.

Here’s what Nvidia’s CEO had to say about it:

Nvidia CEO Jensen Huang has a unique take on what kids should study in school.

While computer science and programming were once the must-have skills, Huang believes that the focus should shift to specific domain knowledge such as biology, chemistry, or finance. pic.twitter.com/ImsKiVvcXK

— Mohit Chaprana (MCR) ?? (@ammohitchaprana) February 22, 2024

Over the course of the last 10 years, 15 years, almost everybody who sits on a stage like this would tell you that it is vital that your children learn computer science. [That] everybody should learn how to program. And in fact, it’s almost exactly the opposite.

It is our job to create computing technology such that nobody has to program and that the programming language is human. Everybody in the world is now a programmer.

This is the miracle of artificial intelligence.

Jensen Huang, Nvidia founder & CEO

Specialise

The future of human work is in specific domain expertise – branches of science, manufacturing, farming, construction, education, engineering and so on.

Understanding the challenges and then using autonomously intelligent technology to help us provide solutions, without us needing to spend months or years developing necessary software by hand.

For most of us it is irrelevant how a particular solution is provided. We’ve learned, or were advised to learn, how to code, because that was the only way humans could interact with computers.

If you wanted a machine to do something, you had to know how to speak to it in its highly technical programming language (or, more typically, a few of them at least).

But the rapid rise of AI is promising to make these skills obsolete for most people other than those designing and maintaining the systems the rest of us will depend on.

It’s already been proven possible, with early experiments indicating that a job that would typically be performed by a team of people over several weeks could be done by AI in just minutes.

We’re not there yet, of course, but we’re on a rapid trajectory that could derail your tech career plans within just a few years.

It makes far more sense to specialise in a particular industry or profession to understand its problems or possible avenues of future improvement so that you know what to ask AI for.

Because, while artificial intelligence can come up with any number of ideas, it can’t really know much about the reality of your work, beyond the information you provide it with. This is why your expertise in the matter is so important so that you can ask the right questions and it can provide you with the best possible answers.

Pick your street

I have to say that the blanket advice of “learning to code” has always felt wrong to me. This is because we all have a very limited amount of time, and if we spend it on one thing, we automatically can’t invest it in something else.

And, realistically, how many of us can be great at programming? (even if we ignore the sheer diversity of what it can actually mean)

Like with human languages, to be good at using any computer language you have to practice it and consistently upgrade your skills with time.

Superficially, “learning to code” can give you some basic understanding of how making computers do things works, but little beyond that unless you are constantly exposed to it. And even within the domain of programming, there are endless possibilities and specialities that you can pursue.

So, “programming” as such is an extremely broad term that is really empty on its own. Programming what? How? Why? For whom? At what scale?

Kids should, perhaps, be taught the fundamentals of computer science, much like they’re taught teaching human biology, even though most of us will never be doctors or scientists. A basic understanding of how computers power the AI-enabled future work is certainly valuable.

But only those who see themselves working as computer engineers should pursue specific skills, with the rest of us being directed towards whatever we’re passionate about and/or predisposed to, as that increases our chances of being really good at it.

So, don’t learn to code — learn to be great at something useful, whatever it may be. AI will always be there to help you out.

The pandemic has greatly accelerated changes in the workplace after millions of people around the world discovered the benefits of working from home. However, not every job can be performed remotely and there are many advantages to showing up at the office, as most people work in teams.

One response to these challenges was a proposal to, instead of working remotely, simply reduce the work week by providing an extra day for personal affairs. This would give employees a greater work-life balance, which may produce productivity benefits, making up for the time reduction.

And while there have been many promising trials of the idea around the world, one of the longest ones has, unfortunately, just ended in failure.

Magyar Telekom, the Hungarian arm of Deutsche Telekom, a German telecommunications giant, which has conducted the test for the past 20 months, has decided to return to a traditional 5-day schedule at the end of February.

The trial of the 4-day, 36-hour week, which covered 300 people across the company of around 5000 employees, began in the summer of 2022. Participants were selected across different departments, including customer service, technical fault repair, technology sales support, and SAP systems.

The goal was to collect information about working efficiency under tighter, more intensive weekly schedules in different roles while providing the same basic salary.

Disappointingly, after more than a year and a half, the drawbacks outweighed the advantages, and the company was returning to business as usual.

Started with a bang…

The program began with a 4-month run in 2022 involving 150 staff across four teams and was initially considered a success:

“Building on international experience, we found that three consecutive days off make work and leisure time more balanced, improve colleagues’ quality of life and enable them to perform more effectively in the workplace with well-coordinated work organisation.”

Chief People Officer of Magyar Telekom Zsuzsanna Friedl, November 2022

Internal surveys suggested that the performance of customer-facing workers improved by 10 per cent, while over 90 per cent of them felt they now had enough time for their personal lives and would prefer to stay on the new schedule in the future.

It all looked like a win-win situation. Encouraged by this promising start, the company doubled the number of participants and extended coverage to more departments throughout 2023.

This is when cracks started showing.

…ended with a whimper

As admitted by the company in its recent announcement, not all employees ended up performing as expected.

In fact, the majority struggled to maintain the same efficiency at work and while an additional day off sounded like a very attractive benefit, adding an hour to each of the four remaining work days couldn’t always be handled by employees due to personal obligations.

Beyond individual performance, the new arrangement also upset the scheduling of collaborative work across the entire organisation, as not all workers could be placed on a shorter week.

This is why, given that:

• not everybody can work only four days
• many people handled it poorly
• scheduling conflicts would make it impossible to employ two different regimes for workers in different departments, who often have to work together

the company has made a decision to conclude the trial and return to a uniform 5-day work week starting in March 2024.

“After a year and a half of testing, it became clear that it would not possible to introduce a solution that can be used uniformly by all employees, due to individual or work process restrictions, and the parallel operation of the 4-day and 5-day work schedule would represent business risks in the long run. In light of this, the company’s management has decided to act responsibly and to close the pilot period of the 4-day working week on February 29, 2024, with continued effort to find a solution that can be undertaken by as many colleagues as possible and generally ensures an improvement in work-life balance.”

Not dead but not for everyone

These results are some of the most important ever recorded, as they reveal that many short trials may provide conclusions too quickly, particularly for large corporations.

After all, the first four months were very promising for Magyar Telekom too. It wasn’t until the run was extended to almost 2 years, involving considerably more people and units of the business, that problems began to manifest themselves.

There is some evidence to suggest that a 4-day work arrangement can work in smaller companies as well as those where flexible or remote work is already practised.

It is certainly much easier to pull it off when your business does not involve daily provision of products or services, or direct interactions with clients/customers.

That’s why if you’re holding out hope for a shorter week at work, not all is lost yet. The idea is certainly not dead but it is clear it’s not for everyone either.

Disclaimer: Opinions expressed below belong solely to the author.

Mark Zuckerberg is clearly upset that Apple can get away with selling a $3500 headset that people around the world are fawning over (even if recognising that not everything about it is perfect), while he’s struggling to make money with his $500 Quest 3, as Reality Labs — Meta’s VR/AR business — is bleeding tens of billions of dollars each year.

Why would anybody buy Apple Vision Pro?

So, as he’s settling into his role of Steve Ballmer of our time, he recorded a short video, claiming that Meta’s Quest 3 is not only a better value than Apple’s Vision Pro but a better product overall, full stop.

In short, according to Mark, you can do a lot more with Meta Quest 3 (play games, watch videos, work etc.); it is also lighter and more comfortable, and the only thing that Apple is delivering for $3000 more is higher resolution.

In other words — why would anybody buy Apple Vision Pro over Quest 3?

Well, before we get to that, let’s see what Microsoft’s Steve Ballmer said about the iPhone when it was announced in 2007:

“500 dollars? Fully subsidized? With a plan? I said that is the most expensive phone in the world. And it doesn’t appeal to business customers because it doesn’t have a keyboard. Which makes it not a very good email machine. […] Right now, we’re selling millions and millions and millions of phones a year.  Apple is selling zero phones a year. In six months, they’ll have the most expensive phone by far ever in the marketplace…”

Steve Ballmer, January 2007

When was the last time you saw a Windows Phone?

They went extinct within a decade of Ballmer’s boastful Apple-bashing, along with the behemoths of Nokia and Blackberry, once ruling the mobile phone market.

The expensive toy without a keyboard, that so many people saw the iPhone as, is today making the most money, even if it doesn’t sell in quantities as large as Android phones do.

Apple is responsible for little over 20 per cent of all shipments but they generate 50 per cent of the global smartphone revenue and a whopping 82 per cent of all profits from the business.

This is what Apple is aiming for with Vision Pro.

Jack of all trades, master of none

Like an oldschool nerd that he is, this is what Zuckerberg is building at Meta — an all-round device that you can use for nearly everything but one that does not stand out in any particular way.

His vision appears to be that one day the Quest devices will be able to do anything you want, that you will never take them off and live in the metaverse — a term that, notably, nobody is using anymore, after it has become public mockery before it ever took off.

Now, this idea may not be all that different from Apple’s ultimate goal, in some distant time in the future.

Perhaps, indeed, one day we’re going to be putting on a set of digital goggles to do anything: gaming, working, watching movies and so on.

What is different, however, is the approach.

Instead of trying to cram in as much as possible into a single product, Apple is focusing on making at least one thing about it really great today. And with this first generation it’s the image quality of its displays, that currently nobody in the IT mainstream can really match.

Unlike Zuckerberg, who tried to convince us that we’re going to live in virtual worlds, meeting with cartoonish avatars of our buddies, Apple recognises — in a spirit true to Steve Jobs — that technology should blend in with what we already do and how we live our lives.

In other words, while Meta started with Virtual Reality (VR), Apple is starting with Augmented Reality (AR). How Zuckerberg doesn’t see this fundamental difference is both perplexing and comical.

For that reason, these devices shouldn’t even be compared. Meta has sold millions of cheap headsets of several generations, mostly for people looking for relatively affordable entertainment in VR.

Apple is selling Vision Pro to those who want to add high-end technology replacements of many of their existing devices, making for a much more minimalist lifestyle that is so fashionable these days.

Just look at the example below:

With AVP you no longer need a 70 inch TV or a massive desktop screen to do meaningful work on. You can just carry your Macbook around and expand both your work space as well as create a virtual media centre for when you watch movies, on demand.

You can append your spaces, walls and furniture with app windows wherever they might be useful, to provide reminders, serve as timers, guide you through cooking, fixing things in the house and so on.

And with high-quality passthrough video, you don’t even need to take the device off as you walk around — and don’t feel like you’re stuck in a blurry 720p render of the world around, like you may using Quest.

Yes, you can do more things with Meta’s device but none of them is really outstanding.

It’s good for gaming, good for entertainment, decent for work — but it’s not great at any of them. This is why people will not pay more than $500 for it, because it’s still just a middle-of-the-road compromise in a plastic case, that most customers will use for fairly brief gaming sessions, little else.

Apple, on the other hand, stated the function of the device clearly: spatial computing.

Adding a new dimension to the technology in your life, projected to your eyes through a single high-quality device.

They don’t overpromise on functions — they merely promise that whatever you do on AVP will blend in with surrounding reality as seamlessly as it is currently possible. That’s it.

All-you-can-eat buffet meets fine dining

And there is place for both. This is another judgement error Zuckerberg is making, probably because he’s so frustrated with the attention Apple is getting despite only entering the new market.

Meta offers a little bit of everything for everybody but Apple, in its trademark way, has created another product that is a status symbol, like the iPhone.

Just like you visit a food court to fill yourself for a few bucks but splurge on a nice restaurant for events that really matter, you don’t buy the Quest because you want the Quest but because you want to use VR. But you buy Vision Pro because you want the Vision Pro.

And while that may appear to be rather vain of us, it’s really good for business.

For that reason Apple isn’t really in competition with Meta, even though Zuck’s ego appears to be bruised. I have no doubt that Quest is going to vastly outsell Vision Pro just like Android phones keep outselling the iPhone.

But it is also clear that Apple is going to make far more money from a considerably smaller customer group, while Meta’s Reality Labs are going to be stuck seeking profitability.

The most mediocre billionaire

Mark Zuckerberg is currently the 4th richest person on the planet, with a net worth of over $166 billion and yet, among fellow Silicon Valley technologists he’s somewhat an aberration.

Despite his enormous success, you’ll struggle to find greatness in anything he does. Most other big names — Musk, Bezos, Gates, Page, Brin, Jobs — have produced multiple ground-breaking products and services, setting new standards for years.

The only thing that Zuck has ever invented was the Facebook timeline, which became the default UI in the social media age, introduced in a time when Myspace required you to visit individual pages of your friends manually.

But ever since it launched 20 years ago, Zuckerberg has only been a follower, typically late to every new trend and regularly failing at innovation.

He built his business empire through acquisitions of companies that successfully did something he struggled with — like Instagram or Whatsapp — often neglecting them somewhat down the road, opening opportunities for other competitors (like TikTok today).

I’m sure most readers will agree with me in observing that there’s nothing “great” about Facebook, which is still plagued by bugs, oppressive moderation policies and endless spam and phishing attempts, that the company appears to do nothing about.

Instagram, similarly, feels stuck in time. It’s 2024 and the web version is barely usable, while nobody at Meta thought to come up with a solution for posting links on the platform, currently still served by clunky 3rd party link-in-bio services.

His venture into VR — and its results — were no different.

Facebook acquired Oculus for $2 billion 10 years ago and this is how the flirt with Virtual Reality started, costing tens of billions more in subsequent years with little to show for the money spent.

The metaverse, that we were supposed to transfer our lives to, is all but dead, despite Mark making a big bet and even renaming the company for it.

Oculus devices are very good for what they offer but they too fall short of greatness, which explains Zuckerberg’s conniption upon seeing Vision Pro threatening to take the customers with the deepest pockets away from him.

But that also makes him a hypocrite, since he tried to approach them with his own upmarket Oculus Pro, launched in late 2022 for not an insignificant amount of $1500.

The device was widely panned in reviews, failed to gain any traction and was effectively discontinued in mid-2023.

That’s why seeing him now touting Oculus 3 as great value vs. Vision Pro is rather amusing considering how badly his attempts at making a premium product bombed just months ago.

He spent a ton of money making a good average product, which by his own murky descriptions has no single clear use case that is well-communicated to buyers. Is there any wonder why it keeps making catastrophic losses?

His problem is lack of commitment to detail, both in conceptual phase and in execution. Everything he does seems like it emerged from a cloud of chaotic ideas, which explains why he can’t grasp the very basic differences between Meta’s headsets and the latest Apple device, leading his entire VR business astray.

Fortunately, other units make enough money to make up for the shortfall, elevating Meta’s stock to new heights recently, which is why it’s unlikely that shareholders will start calling for his head anytime soon.

But it’s equally unlikely that anything that Reality Labs produce will hold its own vs. ruthlessly organised machinery at Apple — one of few companies Zuckerberg can’t even hope to acquire.

Disclaimer: Any opinions expressed below belong solely to the author.

“The era of rendering documents using Photoshop is coming to an end,” announces OnlyFake’s Telegram account. The underground site is offering (or, as of now, used to as it went dark for the time being) realistic images of fake IDs like passports or driving licenses from 26 countries for just $15.

The documents are generated in seconds using whatever credentials and facial pictures you choose (or generate them for you), and placed in surroundings that mimic real spaces an average user would take a picture of their document on – like furniture, bedding, carpets and so on.

In a conversation with 404 Media, who ran this story, OnlyFake’s founder “John Wick,” claimed that the IDs could bypass KYC checks at Binance, Kraken, Bybit, Huobi, Coinbase, OKX and even more traditional, mainstream Revolut.

404 Media itself has managed to pass the verification process on OKX using the driver’s license pictured above, proving that criminals using AI to hide their identities may be more than a step ahead of fintech platforms, with no easy solution to the problem.

Can’t trust your eyes anymore

Even a few years back, some companies requested you take a photo of your ID alongside your actual face or, like Facebook occasionally does, use a short video to prove you’re a human instead.

But AI can fake this today as well, as one company in Hong Kong learned recently, having been scammed through a Zoom call, where all participants except for the targeted individual were AI-generated personas of the company’s personnel (including a UK-based CFO), successfully convincing him to wire out $25 million dollars.

AI has gotten so good so rapidly that humans are struggling to catch up.

Many are advocating (and employing) the use of AI to fight AI fraud, including in security and user/customer identification.

But as AI is getting better very quickly the question arises – how long is it before it can mimic real life perfectly? What then?

It seems that the only real solution may be using biometric data that is unique to a particular person and could be checked against a central database of verified human beings. That, however, would be another security nightmare and a very serious threat should such a database ever get accessed by less-than-savoury characters.

Soon, the choice may very well be between fighting an uneven battle against fake identities or risking upending the lives of real people, should their identities be used for criminal purposes.

It’s that or back to in-person visits in physical locations to do anything that requires confirming our identity.

As we know, criminals have no scruples doing that today, with whatever leaked information they are able to get their hands on.

Following the 404 Media investigation, the OnlyFake site went offline, but it’s clear that doesn’t really change anything. Anybody with as little as a capable PC running an open-source image generation model could train it to produce similar images, with only some fine-tuning required to make them accurately realistic.

The genie is out of the bottle and there’s no putting him back in it.

Disclaimer: Opinions expressed below belong solely to the author.

In a painful fall from grace, the once viral low-cost e-commerce marketplace Wish is to be acquired by Singapore’s Qoo10 for just US$173 million in cash.

Almost exactly 3 years ago, in February 2021, the company’s value peaked at over US$18 billion, following its IPO in December 2020. It was mostly downhill from there.

Following fierce competition from other platforms sourcing impossibly cheap goods from China – like Aliexpress, Shein or recently Temu, among others – and a scandal which revealed that Wish itself ran a fake store, scamming its own users to collect data, the company lost 99% of value in little over 2 years.

“There were unbelievable bargains on “bestdeeal9,” a store hosted on the e-commerce platform Wish, including a $2,700 smart TV being sold for $1 and a gaming computer advertised for $1.30.

But none of the offers were real, and Wish knew it.

The company, an online novelty emporium that had more than $2 billion in sales last year by dangling hard-to-believe discounts, created “bestdeeal9” as an experiment. Listings that had been removed for violating Wish policies were reposted on “bestdeeal9” and used in part to track whether shoppers complained when their orders never arrived.“

The New York Times, July 2022

Q(u)oo Vadis?

So, where is this going? For Wish and its investors, it’s just a way to get back as much money as possible, chiefly from the accumulated net operating losses, worth a total of US$2.7 billion, which can be channelled into a new business (reducing its future tax liabilities).

ContextLogic (CL), Wish’s publicly traded parent, will remain in operation. Only Wish’s assets and liabilities will be sold to Qoo10. CL will effectively become a clean slate with US$2.7 billion in future tax deductions.

Meanwhile, for Qoo10 it’s a deal that considerably increases its international reach, assuming it can monetise it.

The Singapore-based marketplace is present in multiple countries in Southeast Asia, as well as Japan, South Korea, Hong Kong and China.

The bulk of its business, however, is carried through 3 marketplaces: Japan, Korea and Singapore.

According to Similarweb, the Japanese site receives between 12 and 14 million visits per month, followed by 2 million in Korea and 1.5 million in Singapore.

The acquisition of Wish should at least double these figures, opening opportunities for Qoo10 to reach substantially more diverse customer groups worldwide.

Besides that, the Wish brand surely carries some residual value, even if it’s been tainted by the company’s questionable history. A relaunch could once again attract visitors, and there are hundreds of millions around the world who have had some dealings with it in the past.

Under new management, with a new product offering and improved customer service, it may fly yet again. If that goes to plan, then Qoo10 could make a leap from its currently highly localised and relatively small operations onto the global stage.

We can only wish them all the best.

Disclaimer: Unless otherwise stated opinions expressed below belong solely to the author.

Singapore Department of Statistics released its latest Key Household Income Data 2023 two days ago, reporting that the median monthly household income from work (inclusive of employer CPF) increased to S$10,869 last year among resident households (citizens and permanent residents).

Median figure is used as it is the central value among all households — half of Singaporean residents make more and half make less. It is more relevant than a simple average, which is higher but skewed by very wealthy households which make considerably more than most Singaporeans do.

Inflation eroded some of the 7.6 per cent increase in nominal terms, leaving just 2.8 per cent when corrected for the rising prices. Fortunately, the change is still positive.

In money terms, it translates to S$770 more compared to 2022, for a total of S$9240 in additional income from work over the entire year.

It does not include any transfers from the government or other sources of income (such as rental).

Rich keep getting poorer

Somewhat surprisingly, perhaps, it was another year which saw the incomes of the wealthiest Singaporeans decrease in real terms. However, unlike before 2023, the poorest also suffered a decline:

Fortunately it seems to have been just a hiccup in an otherwise positive trend, which has kept incomes of all groups but the richest growing over the past decade:

It’s worth noting how big the gap between the two 5-year sub-periods is for the richest locals.

Despite keeping pace with other income groups between 2013 and 2018, at around 20 to 25 per cent, the wealthiest deciles suffered a significant drop in income growth or even a decline in real incomes among the top 10 per cent of the society.

They are more than 6 per cent worse off than they were all the way back in 2018.

Inequality

For those concerned about inequality, however, this is good news, as the Gini coefficient followed suit and continues its falling, indicating shrinking income inequality levels, further accelerated by government handouts:

S$1500 extra from the government

Speaking of handouts, government schemes aimed at reducing the burden of the transition to higher GST rates for most regular households, boosted the total average received for each household member to S$6371, up by S$512 per person compared to 2022.

Given the household size of a little over three pax, it means that all received an average S$1500 extra in government transfers last year.

In reality, it was even more for most Singaporeans, as the wealthiest pay by far the most in taxes while qualifying for the fewest benefits.

As you can see, residents of the smallest HDB apartments received S$1244 more than in 2022, per person – more than two times the average, while all groups below 5R and Executive Flats still stayed above it.

S$11,000 or more in 2024

Given the slowing inflation, the nominal increase in incomes should also slow down this year, although it is still expected to come in at at least 4 per cent.

This would take the median to around S$11,300 per month, adding another S$5000 annually per each middle-income household.

And while it may not result in a huge bump in the real value of local salaries, it makes Singaporeans relatively wealthier when they travel abroad, given how strong the SGD continues to be against other currencies.

That’s why, despite some struggles with rising living costs (which are, however, a global problem), the Year of the Dragon might be the year of globetrotting, even for an average Singaporean.