On the effects of job automation in the Philippines, drivers of the next crypto bull run, technical talent in the Philippines
Colin’s Mailbag, June 2019
I asked the Philippine startup and crypto communities for mailbag-style questions pertaining to cryptocurrencies, startups, technical talent, and more.
If you have any questions for a future mailbag, feel free to ask them in the comments here or Tweet them to me directly (@Goltra) — thanks!
Which Philippine jobs/industries are most “at risk” due to tech, say, in the next decade? What kind of jobs will disappear/are disappearing? What skills should Filipinos be learning?
(via Gio T. on Twitter)
While job automation (and, *groan*, “the future of work” 🙄) is a high-complexity issue, I think the somewhat naive approach when exploring job automation is to immediately look at what work is currently being done manually (and/or thoughtlessly) and to assume that a robot should replace that chore. While this may ultimately hold true, I think timescale and sequencing are important in this exploration of the issue. Automation won’t simply start with the easiest tasks, it will start with the easiest high-margin tasks — and this distinction is material.
When I first moved to Metro Manila, it radically changed my perspective on labor as a business input. In San Francisco, due to various labor laws and the sky-high cost of living (among many other factors), labor is almost always one of the highest expenses in operating a business, regardless of what type of business. When I got to Metro Manila, I was quickly surprised to see grocery stores replete with so many cashiers (and backup cashiers!), restaurants with incredibly high waiter-to-table ratios, and, by far the largest surprise, actual humans pumping gas at gas stations (!!!). In the U.S., the death of the gas-pumper profession is almost always used as a parable to show the effects of automation.
All of this is to say that the technology already exists to automate many of the low-skill jobs that people might think of as eminently susceptible in the Philippines, but, for better or worse, the unit economics of the labor input don’t necessitate this kind of replacement just yet. Given both the population growth dynamics and urbanization trends in the Philippines, barring any sort of major government intervention, I don’t see the low-end wage economics changing anytime soon. In a strange, perhaps unfair way, the downward price pressure on low-end wages will likely serve to defend those jobs against automation for the foreseeable future — the price-performance of this type of automation technology (self-check cashiers, etc.) will have to move quite a bit further before this replacement decision is much of a widespread issue.
In the U.S., the jobs that are most threatened by automation are done so entirely with an analysis of cost structure. In the face of the movement to more than double the U.S. minimum wage to $15 per hour (i.e., a change the labor input), McDonald’s plans to implement self-serve kiosks in all U.S. locations by 2020. The significant R&D investment by Google (Waymo), Tesla, Uber and others into self-driving automotive technology for commercial trucking is done so because that’s a high wage/high margin business to disrupt with automation.
With this in mind, the obvious high margin, high repetition workflow that comes to mind in the Philippines is the BPO (call center) industry. By nature of its very name, Business Process Outsourcing is inherently process-oriented and the ultimate decision makers in the BPO industry are foreign firms that are already much more accustomed to automating labor inputs.
As it stands, nearly all of the call center support workflow is already scripted and easily broken into decision tree / if-then flows that look almost exactly like what a developer would mock up on a whiteboard and pseudocode prior to writing an algorithm — if you’re dev, a call center script is basically a perfectly written product spec.
At this stage, the primary advantage that humans still have are in comprehension/communication and adaptability, and both of these are domains in which human advantages are quickly being eroded away by advances in Artificial Intelligence.
For comprehension and communication, advances in Natural Language Processing (NLP) and simple things like chatbots will ultimately reduce the natural human unwillingness to interface with a machine (“Alexa, why is my credit card not working?”). It’s not that automated call centers will just suddenly begin to pass the Turing test, it’s more that consumer comfort and experience in interacting with various types of bots (voice assistants, chatbots, concierge services) will simply increase as UX standards solidify in the automated customer support space.
For adaptability, machine learning will ultimately be able to train on the existing set of call data (“this call will be monitored for quality assurance purposes…”) and learn how to capture many of the edge cases that humans would ultimately be forced to deal with. For those few edge cases that can’t be properly addressed, they will just be escalated to the handful of remaining humans in the operation.
As it stands, this innovation is already happening, the International Customer Management Institute estimates that 15% of global call center traffic will be covered via Artificial Intelligence in just 2 years. Locally, the Asian Institute of Management estimates that 800k Philippine call center workers will be replaced by AI in the next 5 years. Intuitively, both these estimates feel low, as professional and academic forecasters are notorious for underestimating the magnitude of winner-take-all technology adoption games (what Taleb might call extremistan effects). At the moment, I am not familiar with what specific actions the Philippine BPO industry is doing to combat this secular change.
Abstracting out a level, the jobs that automation disrupts will not fully be removed — automation isn’t a simple swap, it’s a point of leverage for fewer humans to do the work that previously took many. Some of this automation is already on our shores, but it won’t put us all out of work, it will simply evolve the nature of what work entails.
So what work will remain and what skills should the Filipino of the future focus on? In a world where much of the monotonous work is automated, new forms of digital work — built around weird, previously very poorly monetized things like self-expression, empathy, attention — will arise that can be done anywhere in the world and will be available to anyone who understands how to interact as a digital native.
I’m by no means pretending it’s some simple Tech Utopia where everyone is just a high value software engineer or professional YouTuber, automation is going to lead to some pretty brutal displacements, but population demographics (less than 5% of the Philippines is aged 65 and above) will actually keep the Philippines more insulated from the level of havoc that automation threatens in more developed (older) western countries.
As the Philippine economy becomes increasingly digitized, there will be an entirely new universe of digital professions, gigs, and bounties that can be done from anywhere in the world, many of which we can’t even currently conceive of. As a super-young, digitally literate, high-growth economy, millions of Filipino youths will ultimately learn to navigate this new landscape natively.
What will drive the next crypto bull run?
(via Mike A. on Twitter)
Well… because it took me a bit to actually write this post after the original question was asked… the rally presaged by this question has actually already started 😬.
I asked this question at the tail end of the dreaded #CryptoWinter and now we are solidly in the trappings of #CryptoSpring.
So, what has changed since a couple months ago?
Honestly, attribution for what causes a rally (even in hindsight) is incredibly difficult. That said, here are a handful of the factors that I am personally watching as the market progresses:
1 — Global Macro Forces
The long term crypto bulls (particularly in the Bitcoin ecosystem) will tell you that the single most relevant driver of adoption and value is the asset class proving itself as a competent tool in safeguarding and hedging against global economic uncertainty. Unpacking that a bit, if you believe that tumultuous political events are likely to continue (as opposed to history simply being over), then Bitcoin and other crypto assets are going to continue getting their day in the sun, as people use these tools to protect their wealth.
To that end, prior rallies have been contextualized and often get discussed in conjunction with the uncertainty around major global economic events: the Sovereign Debt Crisis, Cyprus, Brexit, Trump, etc.
Similarly, we are now increasingly seeing Bitcoin and other crypto assets being discussed in the context of the US/China Trade War, in which the uncertainty in the legacy markets has likely been a boon to crypto markets, as people are looking for places outside of the current system to store their wealth.
2 — Continued Institutional Buy-In, Especially at the Infrastructure Layer
The holy grail of market drivers remains to simply have some tiny tiny fraction of the trillions of dollars in institutional investment capital simply buy a little bit of Bitcoin, but before that can happen our industry has needed to dramatically upgrade its infrastructure. If you compare where things were to the last rally, we now have:
- Major institutional exchanges and custody services being created by Fidelity, ICE (NYSE), and others
- Continued institutionalization of and trust in the native crypto exchange players (Coinbase, Binance, Circle, Kraken, BitGo)
- Institutional players like Facebook and JP Morgan launching their own digital currency offerings, which will familiarize consumers and institutions with using digital assets and ultimately serve as an onramp to the more crypto-native assets like Bitcoin and Ethereum
This piece is worthy of multiple blog posts in itself, and many people will question whether each of these developments are a good thing, but as we prepare for the next major bull run, our ecosystem is unequivocally better-equipped to onboard the major global pools of wealth than we have ever been in the past.
3 — BTC Halvening Cycle
Ben Horowitz and the team at a16z have pioneered the concept of “earned secrets” — information you learn as an insider in the industry that gives you an edge versus outsiders — and to this day, the biggest, open “earned secret” in the crypto industry still seems to be “The Halvening.”
The halvening is the period that occurs every 210,000 Bitcoin blocks (about every 4 years) in which the block reward for Bitcoin miners gets cut in half, essentially leading to further deflation in the supply curve and a dramatic impact to the unit economics of mining and other key ecosystem layers. In mid-2020, the BTC block reward will be cut from 12.5 BTC to 6.25 BTC.
While the halvening is common knowledge in even casual crypto circles, it’s an “earned secret” because it remains a phenomenon that outside capital is still completely in the dark about.
Pantera Capital has done meaningful analysis on the Halvening as a cycle driver:
…it seems that Bitcoin’s price has shown patterns with block reward halvings. We have seen a couple of these cycles where the tide begins to shift roughly a year in advance of these dates. Inflection points occurred 376 and 320 days prior to the 2012 and 2016 “halvings”, respectively…
4 — New Mechanisms for Altcoin Creation, Distribution, and Speculative Frenzy
Finally, the other item that always seems to be a driver of the major bull rallies are the mechanisms to create, distribute, and speculate on higher-beta Altcoins.
In 2013, we had an alt season based around clones of Bitcoin.
In 2016, we had the rise of simple smart contract-based ICOs.
Now, in 2019, we are starting to see the new mechanisms for altcoin speculation, in particular the Initial Exchange Offering (IEO). According to Binance Research (who, incidentally, is very well-positioned to play a critical role in the IEO boom):
An Initial Exchange Offering, commonly referred to as an IEO, is a fundraising event that is administered by an exchange. In contrast to an Initial Coin Offering (ICO) where the project team themselves conduct the fundraising, an Initial Exchange Offering means that the fundraising will be conducted on a well-known exchange’s fundraising platform, such as Binance Launchpad, where users can purchase tokens with funds directly from their own exchange wallet.
With the rise of the IEO, we see that many of the most prominent exchange platforms will exert an increasing influence on the altcoin market.
As has happened in each of the prior altcoin frenzies, retail speculators will need to acquire more-established crypto assets like Bitcoin and Ethereum as their fiat-to-crypto on/off ramp in order to send them crypto-only exchanges that allow them to speculate on the higher risk projects, thereby also impacting the prices of those more-prominent assets.
This driver of the equation obviously pertains much more to “animal spirits” than some of the other deliberate infrastructure and macro drivers, but it is the practical reality of our ecosystem.
Interested in learning more about tech talent in PH [especially, devs/education/growing the talent pool]
(via Jonathan C.J. on the Cryptocurrency Philippines Facebook Group)
As someone who had to build a technical team at Philippine startup, this is a topic that’s particularly near to my heart. I am planning to answer it more fully in an upcoming post about the local startup ecosystem, but I can break the issue of local technical talent into 2 major components that need to be better solved:
- Ecosystem Misalignment: Local universities are fundamentally solving for the wrong success metrics (primarily around immediate post-graduation employment) and that impacts everything from what programming languages are most dominant in the Philippine tech ecosystem (not great) to what types of startups talented locals decide to start (linear scaling)
- Talent Retention: Because the Philippines is English-speaking, any competent Filipino dev can pretty successfully use his or her technical skills as an exit ticket to a much higher wage international market — and the proportion of talent outflow versus returning Filipino talent inflow is overwhelming
These factors make building a fully local technical team with the right experience very difficult, but this will hopefully change in the coming years. Looking forward to expanding on that and proposing some solutions in a forthcoming post (tentatively titled: “What’s Holding the Philippine Startup Ecosystem Back” 😎).
