Nine predictions for 2021 from tech to finance. #1 the big tech/SPAC/bitcoin bubble unwinds.
After a momentous 2020, 2021 will represent a pullback in many dimensions such as tech valuations, cloud/eCommerce/telework penetration, and Asian vs. Europe/USA growth. Not because the trends have reverted, but rather because 2020 pushed them forward so dramatically that they will revert slightly to the trend line. Simultaneously, other areas like COVID, Big Tech regulatory action, Cyber, and Europe’s role will see a significant change. Finally, we will have to wait another year for the next iPhone moment and substantive action on our time’s problems.
Three trends will pull back after accelerating in 2020. Not because the underlying trend has stopped, but instead because COVID accelerated them far beyond the trendline.
The Big Bubble unwinds. Tech markets, SPACs, and bitcoin are all seeing unparalleled bull markets. The underlying thesis is very similar for all. The government is printing money like crazy and driving interest rates to zero. Investors need a safe asset against inflation with good returns. Tech is the future, so tech is safe at any price. It also appreciates double digits, so it is too good an investment to pass on. We have seen this story before: real estate (several times), tech (2001), synthetic CDOs (2008), investment trusts (1929), portfolio insurance (1989), Nifty Fifty (the early 1970s), Tulips (the 1630s). The outcome is always the same. Once appreciation stops, investors get second thoughts and start selling. Once prices fall, the asset class is not safe, so the whole investment thesis collapses. Tech, especially SaaS, is undoubtedly very valuable. However, nothing is worth buying at any price. CISCO went public in 1990 at a $224 million valuation. With the internet, it became clear it was going to be a lot more valuable. Markets took it to over $500 billion in 2001, an x2000. Now, 20 years later, it is worth ~$200 billion, an x1000. CISCO was much more valuable, just not that last x2 valuable (small detail for the people who lost 80% of their investments from 2000 to 2002).
Cloud, eCommerce, and telework stall. COVID has turbocharged cloud, eCommerce, and Telework. As things return to normal, there will be a slight pullback in 2021 vs. 2020. Still, the genie is out of the bottle, and we can expect tremendous growth over the next decade, just not next year.
The West grows faster than Asia. Europe and the US have done a terrible job of controlling the pandemic when compared to Asia. Consequently, they have had significant GDP declines. The flip side is that in 2021 growth will seem faster in Europe and the US vs. Asia as they go back to normal. It is a blip. Asia will continue to outperform for the rest of the decade.
Other changes have started in 2020 and will shape 2021, transforming how we live and see the world.
COVID goes away. With vaccines available, COVID will go away in 2021. Things will return to normal quickly. No one was wearing a mask or social distancing in 1920. No one will, at the end of 2021, with the pandemic under control.
Big Tech on the back foot strategically. Big Tech’s reckoning has started. It will take several years, but the stage is set. Regulation and anti-trust action will level the playing field. No single industry or group of companies can be allowed to dominate the world. Governments have managed to avoid it since the railroad robber barons in the 1880s. Big Tech will not be any different. It is still difficult to picture a world in which Big Tech won’t control everything, but it is a matter of time.
Cyber is considered weapons-grade serious. The US government hack through Solar Winds is the equivalent of a digital Pearl Harbor. We can expect a severe reaction across the world. Cyber is the new equivalent of panzer warfare and needs to be taken seriously at the highest government levels.
Europe finds its footing. With Brexit behind it, we can expect Europe to regain the initiative. France and Germany seem aligned to push the union forward. The Digital Markets and Services Acts are great places to start, with strong obligations to create a level playing field in the digital domain. Regaining digital sovereignty, rekindling economic opportunity, and managing the migration crisis should be the priorities.
Two of the things I would very much like to happen won’t happen for another year.
No iPhone moment. Since the tenth anniversary of the iPhone, pundits have been waiting for the next iPhone moment in tech. The next change of platform. Is it AR glasses? Ready Player One-style VR? Apple’s autonomous car? Quantum computing? Cryptokittens? It will happen in the 2020s, just not in 2021. I am intrigued to see what that turns out to be.
No substantive action on climate change, inequality, or cloud democracy. The clock is ticking, but acting is too complicated. Only a real crisis will put our most significant challenges at the center. COVID has shown us that we can tackle them, but trying to play the ostrich and see if they disappear by themselves is too appealing.
2020 has been an exceptional year. COVID was what made it unforgettable, but we will remember it for what it made clear in our collective consciousness. It made evident three fundamental shifts that will shape the next decade. It put to the fore the three do or die challenges we must solve to move forward. It brought the initial fruits of the five impending technology revolutions.
2020 has been undoubtedly the year of COVID. However, in history books, COVID will be the anecdote. The core of 2020 will be about the transition to a new era it marked. A shift with three components:
Three secular trends moving into deployment to shape the next decade: Cloud, Asia, and Data.
Three do or die challenges becoming inescapable to public opinion: Inequality, Climate Change, and Cloud Democracy.
Five impending revolutions delivering on their promise: Biotech, AI, Electric Mobility, Integrated Reality, and Digital Governance.
Three trends to shape the next decade: Cloud, Asia, and Data.
Cloud, Asia, and Data are not new. They have been visible to many for the last decades. COVID has brought them into the open and accelerated them even further. The world of tomorrow will be cloudified, Asia will have regained its traditional place globally, and data will have replaced capital as the scarce resource. These three trends will change the world we live in, much as the industrial revolution, the combustion engine, or the first wave of computing. 2020 is the end of the beginning of the cloud age, and now we start deployment.
Cloud penetration in 2019 was already substantial: smartphone penetration over 3Bn, Cloud share of IT spend at about 12%, and eCommerce penetration at 14% according to Statista and Gartner. COVID has pushed Cloud over the adoption chasm and made it evident to everyone. Smartphones, eCommerce, Cloud infrastructure, and video meetings have continued unmolested through the pandemic while traditional ways of living and working have stopped functioning. 2020 has showcased a new way of operating that is more efficient, environmentally friendly, and resilient. There is no going back, although the transition will be painful for many if not managed appropriately.
Asia’s rise has been long and steady. At thedawn of the Cloud age in 1990, Europe and North America represented ~55% of the world GDP in PPP vs. 27% for Asia Pacific. In 2019, Asia Pacific had already overtaken Europe and North America, at 44% vs. 43%. COVID has taken this further, showing how Asia can outexecute the West. COVID has been a nuisance for Asia Pacific, with very controlled case counts and little economic damage. To a large extent, thanks to its leadership in new technologies like 5G and IoT, aided both by decided investment and openness to the level of intrusiveness they require. For the West, COVID has been disastrous in its human and economic cost. In the West, we have preferred not to deploy technology to track and trace the virus. The difference in approach and results has pushed forward the predicted date of China’s overtaking of the US economy. It has also made companies with a substantial Asian presence much more resilient.
Data is the new scarce factor of production. Datais quickly eclipsing capital, as capital eclipsed land during the industrial revolution. It is what powers internet giants and SaaS companies that are soaring in market value. It is what has allowed Asia to control the virus effectively. It has enabled rapid vaccine development. Capital, meanwhile, is ultra-abundant. COVID has accelerated zero-interest rates, bottomless public debt, and direct subsidies to citizens, which seem here to stay. Investment capital desperately looks for productive uses: Private Equity, technology stocks, SPACs, Bitcoin, or real assets. As a banker friend recently told me, what do you do when your business’s raw material is suddenly worthless?
Three challenges become do or die: Inequality, Climate Change, and Cloud Democracy.
Inequality. Climate Change and the limits of our eighteenth-century democratic apparatus have been a problem for at least a couple of decades. COVID has forced us to confront them at an emotional level. If we don’t solve inequality, we will face populist government takeover and massive, unstoppable migrations. If we don’t solve climate change, we will devastate the planet and the habitats, creating untold economic damage. If we don’t reengineer democracy, we will face digital monopolies, crippling cyberattacks, and manipulated public opinion. These three challenges are interlinked and will accelerate each other if left unchecked. Time is short. The time to act is now. COVID let out a dirty little secret, we have the tools to solve them, and the world won’t collapse if we use them.
Inequality between people regarding access to opportunity, basic needs, and income has been a growing problem for decades. COVID has made this front and center. For the well to do knowledge workers, the pandemic has been a nuisance. Work and life basics could go online. The problems were accessory, wearing a mask, not seeing friends and loved ones so often. For low-income service workers, it has been a complete dislocation. Work stopped, with severe financial consequences, or carried a substantial health risk. Infrastructure wasn’t there to take life online. Cramped living quarters made everything more difficult. Healthcare wasn’t assured in some countries. The K-shaped recovery is threatening to widen the gap even further. Direct subsidies have allowed keeping the gap at a manageable level during COVID; they will need to continue and expand further.
Inequality between countries has existed throughout history. However, income and opportunity gaps between countries are challenging to sustain in a world where information and capital travel freely. Like iron filings in the poles of a magnet, people in developing countries feel the pull to move to developed countries to access opportunity, healthcare, and basic needs. They are willing to risk their lives and endure hardship to do so. The numbers add up for them. Walls and “common migratory policies” are not viable solutions. East and South East Asia have shown economic growth, and opportunity is the only solution.
Al Gore already identified Climate Change as an existential threat in 2006. Since then, it has gotten even more existential. Global temperature, glacier size, storm counts, or almost any other metric shows this. Weather patterns are changing fast, and they could bring untold economic hardship. COVID brought out another Inconvenient Truth. We can stop it, and the world as we know it won’t collapse. We can live with less travel and less pollution if we need to. Let’s do it for our children. We are setting them up for the equivalent of a decades-long pandemic.
Democracy, in its current form, has had an incredible run. The principles of our democratic systems would be recognizable to classical Athenians, and the US founding fathers would be surprised at how little the system has evolved. Age is showing. It is vulnerable to digital monopolies, social media hacking, cyber attacks, and much worse. We need to move to Cloud Democracy. If not, we will face a false dichotomy between Chinese-style authoritarianism or a failing crippled system. The US elections have shown us both the potential of moving the system forward and the perils of not doing so.
Five technology revolutions ready for prime time: Biotech, AI, Electric Mobility, Integrated Reality, and Digital Governance.
COVID has also shown us that we have an embarrassment of riches in the new technologies that are getting ready for prime time. Cloud is already in deployment, and it is a crucial enabler of everything else. At the same time, there is much more coming, and it has the potential of changing our world radically for the better. COVID has allowed us to get some glimpses of this:
Biotech has shown how cloud bioscience can move at a different speed if we are willing to allow it. Tens of effective vaccines in less than a year, remote healthcare, and instant sharing of COVID data have shown us.
AI has had another impressive year of step by step process into areas that seemed beyond reach. Two examples for the year are Alpha-fold’s protein folding and the new biggest model GPT3, with more than 170 billion parameters, that can write as effectively as most humans.
Electric Mobility is experiencing a golden age with Tesla and many other EV makers booming in the stock market and showing they can deliver as batteries and productive processes improve. The last 5% of autonomous driving seems harder to crack, but there is a lot to be improved with what we have.
Integrated Reality, being able to marry physical and digital reality, has played a central role in COVID. First, digitally tracking the real world through IoT, smartphones, and cameras have allowed Asian countries to contact trace COVID effectively. Second, virtual meetings have replaced physical ones enabling knowledge workers to carry on. Third, Facebook has continued developing viable VR with Oculus Quest 2, which is still not there but is close to being an iPhone moment device.
Digital Governance has mostly returned to the collective consciousness through the latest Bitcoin boom (remember 2017?). However, the rise of DeFi is much more central, with more than $14B already locked in lending and derivative contracts through cryptocurrencies, an x15 increase from 2019.
You might not know Roblox, but most 8 to 16-year-olds do. More than 30 million of them log in daily, about 150 million monthly, more than Minecraft or Fortnite. The Roblox S1 reveals a surprisingly complex company, with revenue recognized over two years, and the Apple/Google tax representing its biggest cost category. Can Roblox grow its users substantially tapping the 24+ segment and APAC? Can it achieve tech style profitability even with the robber-barons controlling the railroads? That will determine if the right price is $15 or $100 per share.
What is Roblox?
Roblox is a unique gaming company. It is not a publisher like Activision Blizzard or Electronic Arts because it has only one title, the namesake Roblox which is not a game is more of a platform. However, it is not a platform like Unity, which is targeted at hard-core developers and is behind many mobile and PC games. It is not precisely like Epic Games, which combines platform (Unreal Engine) and game (Fortnite). However, lately, Epic has been trying to turn Fortnite into a kind of user-generated Roblox. It has similarity to Minecraft, especially around customization and user-generated content, but goes far beyond.
According to their S1, Roblox is a platform in the emerging “human co-experience” category, centered around user-generated content, that draws inspiration from gaming, entertainment, social media, and toys, which “some refer to our category as the metaverse.” Having watched my daughters play Roblox assiduously, I agree with their definition (and my daughters do too).
Roblox is a two-sided platform between creators (of games, events, social media, objects, and much more) and gamers who enjoy that content. The Robox platform’s magic is that it is low-code to start (e.g., my ten-year-old has created a zombie smashing game that my five-year-old plays), but can be extended and built into a full-fledged game (e.g., Bloxburg, Adoptme, Jailbreak). Players can customize their avatar and take it to the different games bringing their friends along and making new ones. Simultaneously, a universe of social media stars has emerged around Roblox, creating a community and a shared content. Roblox, in some sense, has similarities to Youtube, Instagram, or Tiktok as it has made it very easy to create and share games for others to play. In contrast, traditional platforms like Unity or Unreal require professional developers and designers.
Roblox makes money by selling its internal currency, Robux, and a premium subscription feature that allows for cheaper Robux purchases. Robux, in turn, power the whole of the Roblox economy. Players use Robux to buy content and power-ups in games and buy Roblox generated, and creator developed objects for avatars. Creators make Robux from these in-game sales, which can are used for their own gaming or converted back into cash. Robux have been selling at about 1 cent, while creators can convert them back at 0,35 cents.
The platforms’ golden virtuous cycle is already alive and kicking. More players come to play the experiences, which brings more creators who can monetize their creativity. Around this, a community of players and influencers create a conversation that makes the latest game (e.g., Piggy’s latest chapter) or experience (e.g., the Bloxy awards 7th edition) relevant to everyone. Roblox look-alikes are trying better graphics (e.g., Crayta, now in Google Stadia), but it doesn’t seem very easy for them to overcome Roblox’s scale advantage. Roblox low-grade graphics make game creation easier and expand the potential device population that can support the games.
Roblox also has a strong focus on safety. The transfer of Roblox is limited to actual transactions to avoid scams or hacking. Interactions in Roblox seem safe, thanks to an extensive safety staff worldwide.
According to the S-1, Roblox has reached a substantial scale already. It has over 30 million Daily Active Users (DAUs) and has close to 20 million “experiences” (games and similar). These DAUs have brought in $1.2B in bookings in the first nine months of 2020, about $55 per DAU and year. Both DAUs and monetization have grown considerably (~x3 in bookings and close to x2 in DAUs). The pandemic has accelerated adoption, but the latest quarter shows the gains seem to have consolidated (8.7 billion hours in Q3 vs. 8.6 billion in Q2).
Roblox spends bookings in the following way. Close to 30% go to pay the “railroad tax” that Google and Apple charge for mobile transactions. The second category is creator payouts, which are now at 17% (over $200M). Roblox claims that its key priority is to accelerate creator payouts to power its golden virtuous cycle. The third is compensation at 14%, with technical infrastructure at 8% and other at 4%. Roblox already has reached a 30% margin on its Robux bookings.
The GAAP picture of all of this is tricky to understand. Roblox recognizes revenue and associated costs over the next 23 months after the purchase. Consequently, revenue lags bookings, and under GAAP, Roblox still recognizes losses. However, its cash dynamic is exceptionally positive, with $345M positive operating cash flow over the last nine months. Roblox doesn’t need the money from the IPO. It is going public to give its shareholders liquidity.
Roblox’s potential depends on its revenue potential and profitability. Let us take profitability first. Roblox is already at 30% of bookings. ~45-50% of its revenues are payments to Google and Apple and payments for creators. We can expect those to stay constant, as Roblox has pledged it will give creators any improvements it achieves with Google and Apple. The other 20-25% of costs are employees, infrastructure, and general expenses. These costs would benefit from improved operational leverage and could be potentially reduced to at least the 15-20% range, if not further. Consequently, we could consider a bull case of 40% margin (45%+15%) and a bear case of staying at 30% (50%+20%).
Estimating topline potential is far trickier. It depends on the user base and revenue per user. Current revenue per user is about $50-55 per DAU per year, far higher in the US & Canada (>$100). However, the cohort curves shared in the S1 show that Roblox is getting much better at monetizing across regions. We could consider 2030 bookings per DAU at $12/month, in line with best practices like WoW subscribers. $12/month/DAU would take the current 5 cents per hour to 15 cents per hour, still very cheap entertainment.
Demographics. Currently, demographics are pretty concentrated. More than half are in the core Roblox user group of 7-13. According to the S-1, Roblox is managing to expand effectively in the 13-24 user group. Roblox’s key challenge is that the 7-24 demographic in medium and high-income countries will stagnate over the next decade. Getting to over 200M average DAUs would require to break into the 24+ demographic to expand their addressable market beyond 1 billion people. Roblox just acquired the assets and critical talent of Imbellus, a gaming developer for learning and evaluation. Imbellus has famously developed Mckinsey’s gamified interviewing first round. The Imbellus acquisition could extend the platform beyond the 24-year range and the pure gaming use case.
Geographically, Roblox has a similar presence in Europe and North America (~65% between the two), with 25% in the Rest of the World and only 15% in APAC. All regions are growing between 60-90%. However, APAC is a relatively underpenetrated region. Roblox is working on this with increased penetration in Korea and a partnership with Tencent in China. Getting to 200M+ DAUs would require an at-scale APAC presence.
Putting together all the pieces in terms of potential would take Roblox to the $25-30B bookings range with approximate cash conversion of those bookings of $10-15B by 2030. This size would be a 15-20x potential growth from today.
Rumors point to a Roblox valuation in the $8 billion range, double its last private market valuation at $4B. If this were to be accurate, it would look like a bargain. With a potential 2020 operating cash flow of over $500M, a 20x multiple would already support a $10B valuation based on current cash flow generation. If you believe Roblox is a platform and not a game and given its current growth rates and competitive moats, this would be a very attractive valuation.
A bull case takes us far beyond this. Starting with the 2030 $10-15B cash flow generation potential and a 15-20x EV multiple, we could be talking about a ~$200B valuation for Roblox in 2030 if everything pans out. Bringing it back to today at 10% cost of capital, we could justify a $50-70B valuation today.
CNBC put forward a middle of the road scenario comparing Roblox to Unity, which is already public and has more than doubled since its stockmarket debut some months ago. It puts Roblox at $37B valuation.
If we assume the offering brings the number of shares to ~600M (about 20% dilution on the current ~500M), the base value would be a $15 share price range, CNBC would be at about $60, while the bull case would take us to $100.
Roblox is relatively unknown in a category that is still not familiar to investors. Its accounting is complex, and there is no clarity on its addressable market. Even relatively sophisticated investors like Jason Calacanis fail to see it as a platform. Consequently, it is difficult to see investors piling on to this one and elevating it to its bull scenario as it has happened with Snowflake. Roblox could be a slow burn that gradually builds value as its incredible growth potential plays out over time.
From the Roblox perspective, this would make it smart to do a small raise, given they don’t require the cash infusion. Enough to create liquidity and free-float, but as little dilution as possible at an arguably low valuation.
It might be tempting to leapfrog directly into AI, 5G, and blockchain. However, if your business isn’t cloudified yet, it will be a recipe for disaster. Cloud might be boring, but it is the place to start. Do the hard work of taking your company to the cloud and only after reach for the stars.
The world is changing at an unprecedented pace. New technologies come out every day. According to the pundits, AI is the future, data is the new oil, and Bitcoin will reach $500k. The tenure in the S&P 500 is shorter than ever, and companies need to transform. Not changing is probably fatal.
Trying to do AI/Blockchain/insert the latest fad before you have moved to the cloud is trying to teach your child to run a marathon before walking. Execution will fail. Credibility will plummet. Talent will be frustrated. Worst of all, cloudification will be delayed. Tie yourself to the mast and focus on getting the basics right first relentlessly, regardless of the siren songs.
The basics are straightforward:
Are your systems in the cloud, or do you still have on-premises data centers with physical servers? Get rid of those first.
Do you have a microservices architecture with APIs, or do you still have your COBOL mainframe monoliths? Get rid of those first.
Are you deploying in agile, or do you still have multi-year waterfall projects that never deploy? Get rid of those first.
Is your front-office multichannel and automated, or do you still have siloed channels with manual customer interactions? Get rid of those first.
Are your back-office and operations zero-touch, or do you still have mostly manual processes? Get rid of those first.
Do you have a data-lake accessible to all through citizen BI tools, or do you still have armies of people processing information through excels and legacy BI with gatekeepers? Get rid of those first.
Do you have an agile organization with multidisciplinary teams focused on delivering customer value with citizen developer capabilities, or do you still have fiefdoms and siloes that depend on a centralized IT for the smallest changes? Get rid of those first.
Do you have everyone’s workplace in the cloud so they can efficiently work from anywhere, or are you full of local servers and storage? Get rid of those first.
Doing the basics right will be a long, painful slog. It will also be a lot less glamorous than 5G edge-enabled AI blockchain-native agents. However, they will also be a lot more useful for the business, the customer, employees, and the bottom line. Check on each of the eight basics with tangible metrics:
# Cloud/virtual servers vs. # physical servers
# microservices with APIs vs. # of monolith applications
# agile teams vs. # waterfall projects
% of zero-touch customer interactions
% of zero-touch back-office operations
% of employees with access to your data lake through citizen BI tools
% of employees with citizen developer capabilities
# of workplace applications or storage locations that are not totally in the cloud
Maybe give yourself an AI project as a reward if you manage to push the metrics forward substantially. Don’t invest heavily in the fancy stuff until you have the basics right. It won’t work.
We have made mistakes and there will be human suffering and economic consequences. However, we can start to glimpse a path to prevail against COVID19. Victory against the virus depends on all of us working together to make it happen.
We are living through incredible times that move very fast. This is the second week of isolation in Madrid and we know that we have at least 4 to go, probably more. Four weeks ago, COVID was a Chinese thing, Three weeks ago, Italy was already in danger. Two weeks ago we went from Women’s Day marches with millions of people and political rallies to full lockdown. Last week, most countries followed into lockdown.
There is still a lot of uncertainty about what will happen and what is the right thing to do. I have changed opinion quickly over the last weeks, so I don’t feel entitled to give lessons. There will be time to review and learn later. Now we have to act and prevail.
There seems to be a consensus about the two problems we need to solve:
Healthcare: uncontrolled expansion and healthcare system collapse. COVID has demonstrated it can overwhelm healthcare systems if unchecked, causing massive spikes in direct and indirect deaths. It has happened in China and Italy, and it is happening in Madrid. Supporting the health system with personnel, equipment, resources, access etc… can mitigate impacts. However, only limiting expansion through lockdown seems to stop a large outbreak, like China modelled and many countries are trying.
Economics: collapse due to lockdowns. COVID threatens to overwhelm the economic system also. The wheel is not designed to stop spinning. Liquidity support and worker and SME support at a massive scale seem the consensus actions. First taken at scale by the UK and then broadly adopted by most countries.
There is a third problem emerging in the global consciousness: avoiding the re-run of uncontrolled expansion. We need a system in place to ensure this until a vaccine is ready in one or two years. We cannot afford the economic consequences of a second lockdown. Luckily Korea and other Asian countries have already pioneered the system, T3. It requires massive testing, tracking and isolating infected patients and tracing contacts of those patients. We have the lockdown period to put this system in place and avoid a re-run, good weather might buy us some more time until the fall.
So, do your part. Whether it is only you or large resources, there is much you can do:
Support healthcare with personnel, equipment, resources or just avoiding it while it is overloaded. Celebrate healthcare workers.
Comply with the lockdown and support enforcement. Help especially those in vulnerable populations. Celebrate security forces.
Contribute to avoid an economic collapse. Help an SME or a hard hit employee (or a million) make it through. Celebrate the ones who sacrifice most in lockdown.
Be prepared for T3. Some few will have to contribute to the technical infrastructure that makes it possible, we will all have to sacrifice a degree of liberty and privacy to make it viable. Celebrate those who create it and all of us for living with it.
As the Italians are saying “Andrà tutto benne”, all will be well. Stay positive and go for it!
I have tried out the Stadia Founder edition. Technically it ticks all the boxes, at least with a top of the line fiber connection. Commercially its current model seems dead on arrival and difficult to resurrect, given Google’s checkered track record with direct to consumer products. Regardless, I believe that Stadia will be seen as the starting point of the new “Age of the Edge”, in which the content you experience is finally decoupled from the hardware you own.
Google Stadia is Google’s cloud gaming service. Cloud gaming is a new mode of gaming in which the game is not executed locally but rather streamed from a remote gaming server that handles all the processes. If this sounds like “Netflix for games”, you are on the right path. However, complexity is substantially higher. While a streaming movie is the same for everyone and doesn’t require server-side processing, a game has a unique state for each player. It also requires substantial interaction between the player and the gaming server, making the controller another additional barrier to consider. Finally, gaming is also much more computing-intensive than a movie, given that the stream needs to be generated algorithmically on the fly, and usually requires a specific type of computing infrastructure: Graphical Processing Units (GPUs).
Even given all these difficulties, cloud gaming has been around for a while, with many companies working towards the “Netflix for games” vision. Tech startups like Spain’s PlayGiga or France’s Shadow have been chasing the dream for more than 5 years and delivering workable technology for about 2 years. Tech players like GPU-giant Nvidia have also pushed forward for this category to emerge with pilot programs. Microsoft and Sony, the console leaders who have the most to lose have been ambivalent. Microsoft’s XBox Game pass for PC Beta shows what they could do, but they have fallen short of a TV compatible solution that would cannibalize their core console business. Steam, the behemoth of PC game distribution, has also been trying out options. And publishers such as Origin and Blizzard are moving towards Direct-to-Consumer.
I have tried many of these and they all work to a certain degree. Based on my experience I have come to think there are three key elements for a successful “Netflix for games” on TV:
Streaming technology. Getting a virtualization technology that allows effective streaming of gaming without any perceived lag.
User interface. Finding a controller option that allows managing the service and playing comfortably and at a cost-effective price.
Commercial and business model. Providing good value for money in a package that takes care of both technology and content.
We will now look at Stadia on these three elements
Streaming Technology: Very Good
Stadia’s streaming technology works very well. This is not really a surprise, as the different game streaming solutions I have tried over time have worked and Google is a company known for superb engineering. Of course, I am lucky to live in one of the countries with the best fixed connectivity in the world and it shows. With my 600Mb low latency fiber connection in a major Spanish city, the experience on the Chromecast Ultra is perfect. You can’t tell that the game is executing somewhere in the Cloud. Destiny 2 shows extremely impressive graphics and no lag is experienced. I assume Google has a node very close to me so latency is probably in the 20ms range and it is not really executing in a remote Cloud, but rather fairly close.
Several of the reviews online have criticized Stadia for not being “real 4K” but just something very close. While this might be the case, I believe Stadia is good enough for most gamers and TV screens. Those who really care about the perfect image will stay with local hardware for some time anyway. Some reviewers have complained of an “extremely hot” Chromecast ultra, a problem I haven’t experienced yet. Google has also failed to deliver at launch many of the innovative functionalities that were promised such as “Crowd Play”, “Stream Connect” or sharing save data between accounts.
Things don’t work as well on other devices though. I have a low-end PC and there Stadia works in fits and starts on the Chrome browser. It probably has to do with Chrome itself and the Wifi interface getting overloaded but clearly the technology is not ready yet for any hardware. The Chromecast ultra at 79€ is very cheap for a console equivalent, although fairly expensive for an entry-level set-top-box or smart TV (equivalent to 30-40€).
For more remote locations and higher requirements, such as VR, I can imagine needing to deploy Edge computing solutions to lower latency further (to sub 10 ms). This shouldn’t be a problem as this will be available from telco operators over the next years.
So overall, the technology looks very solid and hopefully will progress to Excellent as the service is trialled at scale and optimized. 4K and 8K will be fully taken care of. Connectivity and latency shouldn’t be the problem as fiber becomes universal. Entry-level hardware will be eventually addressed.
User interface: Very Good
The user interface has been a very positive surprise for me. It took under 10 minutes to set up the whole of Stadia on my TV and my children are able to use it without help. The controller is rechargeable and based on WiFi, meaning it has fewer glitches than the typical Bluetooth alternatives. Getting Stadia on an off with the controller shows the power of Chromecast which natively communicates with the controller. Google has shown us how the TV gaming experience should work, and it is a substantial improvement on previous interfaces I have tried out.
Of course, there is a problem. The Stadia controller is $69, a cool 3.5x compared to your typical standard wireless gaming controllers, which don’t work with Chromecast Ultra. So a four-player game around the TV starts at over $350, which puts you in console territory anyway. A proprietary controller also represents a risky bet on Google’s direct to consumer efforts which have a chequered history.
Hopefully, Stadia will open up Chromecast compatibility to other standard controllers allowing to use one Stadia master controller along with cheaper controller alternatives. This would lower another of the important entry barriers for Cloud Gaming.
Commercial and Business model: Very Poor
The technology and the user interface have their blind spots, but are mostly ok. However, the service dies in terms of its commercial and business model. Google has a limited catalogue of games, around 40 compared to Microsoft Game Pass PC with 100. Within those, it has only a limited number of top titles. On top of that, and here comes the real problem, Google expects you to buy the titles at the full retail price just for Stadia.
A quick introduction to how the game industry works. There are literally tens of thousands of computer games, with Steam, the broadest catalogue service, having over 30.000 games. All these games are usually sold on a pay-at-purchase model, with prices very skewed between top of the line so-called AAA games (40-60€) and the rest of the catalogue (10-25€). Some companies have tried out a subscription model (most notably Blizzard with World of Warcraft) and in-game purchases (for example Epic Games’ Fortnite). Purchases are typically for the game on just one platform (e.g. PC or console).
Games have different behaviour from movies in terms of usage. While you typically just watch a movie or series once, a game can be played for a long time. Gaming studios typically aspire for “replayability”, with really successful titles like World of Warcraft or Fornite achieving hundreds or even thousands of gameplay hours. So game consumption is much more concentrated than video consumption. Players want “their titles” to be on the platform.
Google’s Stadia follows traditional industry norms, so it falls flat on a number of issues:
Extremely limited catalogue. 40 titles compared to the 30.000+ games out there. Of course, if you are a Destiny 2 fan you will have your game, but they only address a very limited set of gamers. The catalogue can grow over time, but it will depend on Google’s capacity to build win-win relationships with publishers
Buy for Stadia only. You will have to buy titles just for Stadia and you will be locked in, if Google is not able to develop it. Google is a newcomer into the gaming category and has a history of dropping or freezing things that don’t work. You might be left with hundreds of euros of purchases that only work on a dead service. It would be like having a collection of Laserdiscs or Betamax.
No way to try out games. There is no way to sample games, so your only option is to spend 20-60€ directly in a game and see how it works in the platform and with the controller. This compares with services like Microsoft Game Pass PC which allows you to sample 100 games.
Taking all of this into account it would take a radical shift from Google to be successful with Stadia. It also plays to Google’s traditional weaknesses like partnering with others on an equal footing or creating attractive consumer proposition. It seems unlikely Google will be able to pull it off. Google Stadia’s technical infrastructure, however, might end up being an interesting proposition through Google Cloud. Game publishers need an infrastructure for the future and building it is not their core expertise, so using Google Cloud might be a viable option.
Software is eating the world. The digital transformation has come to all sectors and Telco is already in for the second round. Telcos have significant opportunities to capture across the P&L. Digital Services both for the B2B and B2C segments can boost the top line taking advantage of unique market access. Digital Processes can transform the way telcos execute every process, increasing efficiency and improving the customer experience at the same time. Digital Networks will transform the asset base of telcos, allowing to simplify and improve efficiency. All of this will require some scale, and specially industry collaboration.
Digital Services: optimizing the topline
Consumers and businesses are faced with an exploding array of digital services. To take the two clearest examples. In B2C video offerings are multiplying. First was Pay TV, then came Netflix, then HBO and Hulu, and now everyone is going Direct-to-consumer. We will soon have Disney+, HBO-all-access, CBS-all-access, Apple+, sports specialists like DAZN, and many more. In B2B cloud offerings, we used to have AWS, now Microsoft and Google, as well as Alibaba and Tencent, are rapidly joining the party. As well as literally thousands of SaaS products for different processes or verticals.
This cannot hold. History and psychology show that humans like choice, but not so much choice. Customers don’t want to keep track of which of the tens of services they will need to see their favourite show. So there is an opportunity to rebundle digitally access to content and other digital services. Customers need someone that can bargain for them against the increasingly powerful global digital players. Digital services, even leaders like Netflix or AWS, need a distribution edge over the increasingly crowded field of rivals.
Thus the opportunity for telcos to be the distributors of choice of digital services is already there. The entry of new OTT services created the opportunity for partnerships with Netflix and the like, that many telcos are already leveraging. Much like with physical retailers there is also an opportunity to white label for low-cost with acceptable quality. Reliance Jio in India, for example, has created a suite of basic digital apps it bundles with its services for first-time users of its basic line of sub $100 smartphones.
We shouldn’t only think of consumers, the same opportunity is there for businesses large and small. Companies are increasingly focused on their core competence, not caring to develop other skills that someone can provide from the outside. Finding the right suppliers and integrating the solutions in domains as complex as software-defined communications, internet-of-things, cloud, security, Big Data and new emerging fields like RPA is a daunting task for most companies. Thus telcos with a strong B2B presence are increasingly having an opportunity to extend their core connectivity offering in these areas.
Of course, being a digital services distributor will require new skills. Customers expect the same level of ease for purchasing and cancelling they get from digital services themselves. They also expect a seamless integrated experience with the same customer experience they get from any of the services directly. In the B2B segment, they expect professional services to handhold them through integration and change management. The good news is that these improvements are the same ones required for improving the telco core services consumer experience or B2B value delivery.
Financially the profile of this new digital distributor and integrator business will be very different. While the traditional telco business requires high EBITDA margins to compensate for the high CAPEX and high capital employed, the new distribution business has low capital employed and low CAPEX. Much like Amazon has done with its core Amazon e-commerce business (low capital employed, low margin) and its AWS division (high capital employed, high margin), telcos will have to learn to manage and make visible for markets two very different businesses.
Digital processes: improved efficiency and customer experience
The opportunity in digitizing telco processes is also immense. It is often said in the industry that “the cost optimization opportunity is endless”, and digitization makes the adage true for another decade. Not only that, but the customer experience opportunity is also endless, and digitization will also bring the opportunity to take it to a level far beyond what we now have. For the first time, cost efficiency and customer experience will improve in tandem instead of representing a trade-off.
It is seldom appreciated how complex telco services are from an operational perspective. We expect a global network of real-time connected infrastructure to interact seamlessly with billions of connected devices that move around. We also expect the result of those interactions to be correctly billed in real-time to billions of customers worldwide. The challenge of moving a call or a data session from one tower to the next without dropping. The intricacies of processing the millions of call and data session records that go into a single bill. It is amazing that operators are able to give the service level they give with only tens of thousands of employees.
Digitalization will bring the opportunity to take this miracle even further and at an even lower cost. A set of technologies are already being deployed into production that will transform each of the telco process domains:
Digital front-end. The telco customer interaction layer is one of the sore spots both from the cost and customer experience side. Digital channels and natural language processing have the opportunity to transform this completely. We are already seeing an ever-increasing share of service interactions being digitized and managed with digital agents. This is much cheaper and also gives a much better customer experience. Call-centre agents and shops will be still crucial for complex customer interactions, and they can be enhanced through digital tools that allow them to focus on the customer, instead of being “human data processors”. Dedicating humans to empathy and doing all the data processing and standard tasks through digital will create a new front end that is much cheaper and has much happier customers.
Eliminating the back-offices. Telco IT systems are complex, and a surprising number of a telco staff is now focused on making customer interactions “travel” through this complex and multilayered maze. Each back-office interaction is, first, a cost item and a failure of the system, and, next, an opportunity for mistakes and dismal customer experience. New tools like Robotic Process Automation (RPA) and cognitive machine learning systems are allowing to eliminate the need for human intervention in the back-office. Reducing cost and eliminating extremely frustrating mistakes and problems for customers.
Personalizing and adapting the offer. Next in the stack is the value proposition for customers. Historically telcos have taken a “shotgun approach” with the same offer and communication for everyone regardless of need and situation. Big data analytics and artificial intelligence are allowing to change this. The offering can be personalized for each customer according to situation and needs. Moving interactions from frustrating to delightful. And allowing to discover the ever-widening set of digital services a telco will offer.
Digital field operations. Another often unknown side of telcos is the army of engineers and field workers they employ. They might be installing a customer connection, sourcing and sending handsets to customers, repairing a base station or laying new fibre to upgrade connectivity. This complexly choreographed daily dance is another miracle, but also a source of cost and customer unhappiness. It also represents the opportunity to apply new technologies like Internet of things, augmented reality and Blockchain. Telco operations are being transformed with connected vans dispatched with artificial intelligence, blockchain-based supply chains and enhanced technicians that can access the top knowledge through AR to make a field repair. This will reduce cost and at the same time delight customers through seemingly “magic” service.
APIfied architectures. Finally, the telco IT systems themselves are being transformed. Again, not widely known, telcos are the most IT-intensive industry out there along with Technology and Finance. The need to give real-time service and billing to millions of customers has made IT indispensable from at least the 80s. This means that most telcos have layer upon layer of systems, ranging from mainframes to microservices. This makes connecting with those systems extremely challenging, which has been a traditional barrier for collaborating with third parties and improving the front-end. That is where APIfied architectures come to save the day. They are allowing telcos to isolate partners and the front-end from the byzantine complexity of legacy systems. While the direct impact on cost and customer experience will be difficult to trace, this will be the enabling layer for all the rest of the transformations mentioned in this section.
Digital networks: a programmable and open source asset base
Finally, there is the network. The real deal for an engineer like myself. Again, the telco network is a marvel of human engineering. The global telco network might be the most complex system on Earth at this point, even surpassing a human brain. We take it for granted that people in rural Nigeria can call someone in London in seconds and it will work seamlessly. We know that if we get mobile connectivity or Wifi anywhere we will be able to access the same digital services and world wide web we are used to. Making this happen has unspeakable complexity and layer upon layer of equipment interacting in milliseconds.
Digital allows transforming these networks into a new software-based paradigm. Creating very attractive new opportunities for telcos to move to a more liquid asset base, that serves customers much more flexibly at more attractive costs. It also has the potential to future proof the network, allowing to facilitate network evolution significantly. In a way, there has never been so much clarity about the future of telco networks for the next decades, and at the same time, there are plenty of opportunities.
The best way to understand this potential revolution is to take each layer of the telco network at a time. Looking at transformative technologies and their implications:
Fixed Access. The future technology of fixed access is fiber+Wifi. Fiber has incredible potential and upgradability in terms of bandwidth. Its passive optical nature makes it environmentally and operationally far superior to other active electrical alternatives (e.g. copper, cable). It also has great latency characteristics that will be increasingly important over time. A full optical IP access network also requires a fraction of the equipment that traditional technologies do, liberating substantial space, CAPEX and power consumption in local exchanges. This equipment is also increasingly open, allowing for many companies to design it instead of being locked into proprietary standards from large technology providers.
Mobile Access. 5G is the future of mobile access and will represent a revolution as important as the move from circuit-switched to packet-switched technologies. 5G will bring greater speeds, far greater number of devices and lower latencies. It will allow the creation of network slices, softwarizing mobile access and giving it the flexibility of software. 5G is still in its initial stages, and for at least 2-3 years its usefulness will be limited. However, once the final standard is in place we can expect a true revolution over the next decade. This revolution will also include open source equipment which breaks the hardware-software integration pushed by large technology providers in the previous generations.
Edge Computing. Fiber and 5G are a perfect match to create an extremely high bandwidth network that will enable the next generation of rich media content: Augmented Reality, Virtual Reality, Digital Twins, AI models, HD streaming and much more. These new applications will require much lower latencies in the millisecond range. For this to work computing and storage will need to move from the datacenter to the network. Telcos have the assets to do this and play a new role in the cloud.
Homes. The home is also part of the network, and increasingly turning into the bottleneck. Ensuring adequate WiFi in the home is becoming increasingly important for users to take advantage of the full advantage of the network. Making that WiFi work with the rest of the network and with 5G network slices while keeping users secure will differentiate connectivity for users even further.
Businesses. Businesses are also seeing a radical transformation on their networks and premises. To begin with, the new technologies are allowing software-defined virtual private networks that take advantage of fiber and 5G, while providing maximum flexibility in their configuration and deployment. We will see a move from a new private network point taking weeks or months to set up, to it being available in seconds, as soon as the connectivity is ready. Next, connectivity in corporation HQs, factories and other large campuses is becoming the limiting factor. 5G and WiFi 6 will bring a new way to deploy private campuses that will make connectivity work transparently and securely in any setting.
Metro and core network. Finally, hidden from view, we have the metro and core networks. These are the network elements that make the magic happen. Softwarization will transform the complex labyrinth of proprietary equipment that now powers a network into a mesh of transmission, compute and storage infrastructure that seamlessly executes the core and network virtual network functions were they are needed. They will create a more resilient, flexible and much cheaper network
Some scale and plenty of collaboration
To conclude, these three opportunities bring a perennial question to the sector. Given that software is global, wouldn’t it make sense to have global companies take advantage of these opportunities? Will a large global telco more quickly take advantage of these opportunities or will small organizations be equally able to leverage them? Let’s take them one at a time:
Digital Services at the basic level will be about partnerships with large and small digital service providers. Here scale will be clearly helpful for being on the radar and negotiating better. It won’t be necessary to be global, but it will be helpful to be substantial enough to be in the digital service providers priority list. This means scale in terms of high-value customers and ability to cover certain regions. There is also an advanced level, in which operators can create professional services organizations to deploy digital services in B2B, or advanced app ecosystems to deploy a combination of branded and white-label services. Here scale will be more important, both local scale and also a minimum scale to create the capabilities to work with customers. Scale can also be complemented with collaboration as smaller players join larger ones in alliances to strike partnerships with digital players or service organizations to provide professional services.
Digital processes will be about applying technologies to current operations. These technologies are mostly common to all industries and as such professional services organizations can support companies in adopting them. Telcos will be some of the most advanced users in some of them like natural language processing, big data or RPA, so if they have the scale they could transfer best practices quicker or even use that position to leverage digital services. However, digital processes themselves won’t require substantial scale. There might be the opportunity to transfer processes and technology from one market to another, but this will be limited to new entrants or low-complexity operations (e.g. no-frills MVNOs) which don’t have substantial IT complexity legacy.
Digital network is very much a telco undertaking and as such will require telco global scale to happen especially in terms of the open access, open network functions and edge computing. Telcos have started to create their own organizations to make it possible, such as the ORAN Alliance in which most major carriers have pooled their efforts to create a global standard for open radio access. Of course, Digital Network does require a critical mass of software talent in the networking organization, which could be challenging for smaller players.
So overall the digital opportunities for telco will require a certain minimum scale to fully take advantage of, but don’t seem to be poised to trigger a wave of industry consolidation towards greater scale. At the same time, they will require much more industry collaboration. At a global level to standardize the open digital network to capture the opportunity in each of the domains, and especially in Edge computing. And at a more local level as some of the larger companies can support smaller ones in their relationships with digital service providers and in building professional services organizations for B2B and app ecosystems for B2C.
When a twentieth-century Jesuit priest, the prophet of the Singularity and Elon Musk reach similar conclusions there is probably something to it. Each reached a similar and startling conclusion. Humans will be superseded by something greater than them in a relatively short time. In this moment of Brexit, Trump, trade-wars, impending recessions and global warming it sounds crazy. However, given the three got to the same place from three different angles it is something worth examining.
The Ironman route: Elon Musk
Elon Musk is famous for a lot of things, among them world-changing startups and unfiltered comments. His most unfiltered comments are about the future of humanity. Musk is deeply worried about artificial intelligence. He rightly states that we are very slow compared to silicon-based artificial intelligence. Consequently, AI will quickly make us irrelevant when it comes. He uses the Fermi paradox (the paradox that there are no space aliens contacting us even if there are billions of stars and planets) as a proof point that something quite dramatic seems to happen in evolution before galactic space travel. He hopes our new AI overlords will be kind and is trying to at least join the party by creating Neuralink, a way for the human brain to interface directly with technology. He even thinks that all of this might already have happened and that we might be living in a Matrix-like simulation controlled by such overlords.
Musk might sound weird and disjointed when he tries to put together his thoughts (Jack Ma and Elon Musk on AI). However, his logic is impeccable, as usual. If you believe that AI can develop, silicon is a much faster substrate than a carbon brain for processing. So his notion that life might be a “bootloader” for true consciousness on silicon might not be as far-fetched as it seems. That is why Musk believes that AI is the most important topic confronting humanity right now.
Of course, AI might also take much longer than expected. So Musk is trying to save the planet through EVs with Tesla and make humanity multi-planetary through SpaceX. Just in case we manage to destroy ourselves before we finish our bootloader work.
The Singularitarian route: Ray Kurzweil
Ray Kurzweil is also quite a character. A distinguished scientist and inventor, he has successfully created a multibillion-dollar corporation (Nuance) and is now the head of engineering of Google, the church of engineers. This makes it particularly remarkable that he has been able to create a completely new ideology (some might even say religion) in his spare time, Singularitarianism.
Singularitarinism is the belief that a singularity in evolution is near. This singularity will come once AI is able to design more AI making itself progressively better and faster at increasing speeds. Through this virtuous cycle, it will achieve escape velocity from human time scales and creates total change in a single moment in time. Like a black hole (a physical singularity), we cannot predict what is on the other side of the Singularity, so we are left guessing. The key engine behind the Singularity is Moore’s law. It states that computer power roughly doubles for the same price every 18 months. This has created an exponential explosion of information processing capacity. Kurzweil also believes that humans will be immortal relatively soon thanks to technology and that we will be able to upload our brains into the cloud, echoing Musk’s desire to join the party.
Again, Kurzweil’s thoughts seem otherworldly and impractical at first glance. However, like Musk (who seems at least slightly indebted intellectually to Kurzweil), his logic is impeccable. Change has been accelerating almost since the universe was created (stars in billions of years, minerals in hundreds of millions, life in tens of millions, humans in millions, civilization in thousands and technology in tens of years). Moore’s law is just the latest stage in this change. If we project the trend line of change acceleration we should be getting to the Singularity in the 21st century.
Of course, something could change or delay it. So Kurzweil, like Musk, is preparing for the worst. In his case trying to extend his life as much as he can with a variety of tactics which he covers in his book, Transcend: Nine Steps to Living Well Forever.
The Christian route: Teilhard du Chardin
Father Teilhard du Chardin was one of the most disruptive Christian thinkers of the 20th century. A French Jesuit palaeontologist who worked in China, South East Asia, USA and Europe, he represented globalization before globalization was a buzzword. He had a unique combination of influences that led him to unique ideas. Not surprisingly these ideas were not welcome by the church and were only published after his death. The posthumous publication of his work made a big splash in the 1960-80s and he even inspired an award-winning series of science fiction books by Dan Simmons (Hyperion Cantos). Today he is much less known, but his thought resonates more than ever.
Teilhard’s view was that the world was moving towards an Omega, the ultimate goal. A reversal of the traditional Aristotelian view of the world coming from a first mover, an Alpha. He saw evolution as a palaeontologist from atoms to molecules to minerals to life and finally to consciousness with human beings at the apex. “Cefalization”, the increase of consciousness in living beings was the most important trend in evolution for him.
He projected the path ahead as an observer of early globalization. Humankind according to him was converging and being compressed into a smaller and smaller world. This compression would eventually lead to a post-human stage in which consciousness was compressed into the Omega point. He believed in the Omega point as a Christian. The endpoint was Christ and God, the divine parusia, the second coming that had been prophecized in the bible. He put everything forward by combining the prevailing creeds of his time. Religion moving humans upward to spiritually and grace. Marxism and Capitalism moving humans onward to material progress. Teilhard said a new religion was needed for our age that goes upward and onward at the same time. Moving onward towards material prosperity to move upward spiritually and towards God.
Teilhard’s vision of evolution has been confirmed over the ensuing decades. His vision of human compression has proved prescient and materialized even beyond what he could have expected. He died before the start of the electronics revolution, so he didn’t have AI and Moore’s law in his toolbox. This makes his posthumanism very much AI-free, even if he identified “machinism” and “scientifism” as two key trends along with globalization.
The three routes put together: God at the end of evolution
The language of the three thinkers is very different, but they share a remarkably similar story. Evolution gradually leads up to life and consciousness, and eventually AI. Consciousness accelerates and compresses on Earth in an accelerated fashion. Eventually, consciousness compresses into a single point, the Superintelligence, the Singularity or Christ. This process is close to completion, continuous and unstoppable. At the end of the process, there is something so much beyond human consciousness and capabilities that we are at a loss to explain it or understand it. Something that could be said to have all the characteristics of God.
The three visions together form a compelling whole. We need something new to believe in. It can be Teilhard’s new religion, onwards with technological and material progress to move upwards towards God. This gives meaning to the Singularity’s disquieting unknowableness and to Musk’s biological “bootloader”. Kurzweil’s focus on accelerating technological progress through information processing technology gives us the path from which to move onwards, now towards a greater goal than only technology. Finally, Musk puts humans back into the picture. Those who wish can try to participate directly in the parusia through Neuralink, while others can participate indirectly through religion.
We will see what the next decades bring. It will be exciting for sure, but if it is the Omega point we are really lucky to live at this point in time (at least for those of us who are very curious). It is also remarkable how technological and scientific arguments can apparently support an “end of the world” narrative, so popular in many religious settings. I myself see the logic, but I am sceptic enough that I will continue to live my life as if the Singularity was not going to happen. However, I can understand the curiosity that leads both Kurzweil and Musk to try to extend their lives and try to have a front-row seat in whatever happens. Whatever progress brings, even if we just hit a wall and stop moving onwards, it will be extremely interesting to see.
We are living through disconcerting times. Negative interest rates for bonds while hedge funds, venture capital firms and big technology players return double digits. Stagnating wages for the middle class, at the same time as engineers are hired with six-figure salaries just out of school, and companies claim they can’t find workers to fill positions. Polarization in land values, with San Francisco, London or Hong Kong becoming really expensive, while most areas lose value. What is happening? Is this just a phase?
I will argue this is the new normal, and we need to adapt to it. Knowledge has become a key factor of production and it heavily affects the returns of Capital, Labor and Land. This has been a long time in the making. Peter Drucker already used the term knowledge worker sixty years ago in The Landmarks of Tomorrow (1959) and knowledge economy fifty years ago in The Age of Discontinuity (1969). As Carlota Perez claims in her Technology Revolutions and Financial Capital, the full maturation of both ideas has taken half a century.
Let’s explore how each factor of production has been affected and what are the returns and dynamics of Knowledge itself. Along the way, we will mention Attention, a new type of Labor that is emerging as a factor of production.
Capital: so abundant as to be free
We are facing an apparent contradiction. Bond yields have gone negative for almost half of all bonds issued. My own country, Spain, and our neighbour, Portugal, have almost zero rates for its 10-year bonds. Ten years ago they were facing default and intervention. Danish banks have started offering negative interest mortgages and charging for deposits. At the same time, returns to Capital are higher than ever since the 1920s and they have compressed returns to Labor to all-time lows. Some VCs, PEs and hedge funds create outsize returns for their investors. The large tech monopolies (Amazon, Microsoft, Google, Facebook and Apple) have all averaged huge total returns that have made them the most valuable companies ever. New software unicorns are minted every week.
To reconcile both extremes we have to differentiate based on how much Knowledge is injected with the Capital.
Plain Capital. Most Capital is deployed requiring little or no Knowledge. Anyone can buy a German bond, invest in a mortgage backed by a good property or deposit cash in a bank that is insured by the government. It has no risk, so it requires no Knowledge. There is more Capital without Knowledge than we can employ, so the new normal is negative interest rates for relatively riskless Capital.
Knowledge Capital. Some Capital is deployed requiring a very high amount of Knowledge to make it productive. Venture Capital investing, advanced Hedge Fund strategies, high Knowledge companies or successful Private Equity investing, all require high stocks of Knowledge. Thus they can earn very high returns. At the same time, they have very high dispersion. Some Knowledge is right and other is wrong. The challenge is how to differentiate, “good Knowledge” (e.g. Warren Buffet) from “bad Knowledge” (e.g. Bernie Madoff) from the outside and be able to pick the winners.
This new world in which Plain Capital by itself earns negative returns and Knowledge Capital can earn outsized returns breeds inequality. Silicon Valley VCs, New York hedge funds and London PEs can all make a killing. Large pools of capital, like wealthy individuals, family offices or endowments, have access to Knowledge Capital investment opportunities. But the average saver is mostly limited to Plain Capital or to trying to pick Knowledge Capital from limited sources. At the same time, it challenges our financial system which is based on positive interest rates for Plain Capital.
Labor: Knowledge winners and Plain losers
In the Labor front, we are facing a similar situation to Capital. On the one side, life is tough for Labor. Labor share of GDP has been dropping precipitously. Most workers are seeing their jobs automated, shifted off-shore or face so much competition that their wages are deflating. On the other hand, we have a very tight Labor market with firms claiming they have millions of unfilled specialist jobs and that the war for talent is more intense than ever. Engineers and data scientists in Silicon Valley command astronomical starting salaries while clerks and factory workers everywhere are facing wage deflation.
Again, to understand this phenomenon we need to look at the impact of Knowledge on Labor. We will explore the different types of Labor based on how much value they add:
Attention Labor. A new type of Labor is starting to emerge. Attention is very important for companies like advertisers or software developers. It is used to generate revenue by selling advertisement of products and services, and also to generate information that improves products or feeds AI algorithms. Attention Labor is still very low value at an individual level (maybe in the hundreds of dollars per year in the developed world), however, it has become very important in aggregate. The free Attention Labor Facebook or Google collect from the usage of their products is what constitutes a very important component of their competitive advantage. We pay Facebook, Google, TV channels, and other “free” services with almost 12 hours of media consumption and smartphone use per day, our Attention Labor.
Plain Labor. Most Labor requires very little widely available Knowledge. Jobs that require high school level or even basic university-level skills are in this category. Inmigration and the ability to globalize work has made Plain Labor widely available at very low prices. The educated billions of China and India have competed away wages from the developed world, and Africa stands ready to step in whenever Asia is exhausted. If globalization was not enough, we are seeing the rise of the Digital Worker that automates the work of Plain Labor. Under these twin pressures, the wages for Plain Labor are plummeting creating inequality and social disruption. Plain Labor can expect their children to have less economic opportunity than they did if they live in the developed world.
Knowledge Labor. Some Labor is combined with substantial Knowledge. Thus it manages to keep its value and its scarcity allows it to command substantial wages. These are the software engineers, surgeons, M&A bankers, salespeople, coaches, consultants, product managers, data scientists, blockchain experts and many others. They represent the millions of jobs that companies cannot recruit for. However, the advantage is fleeting, these workers are only as valuable as the scarcity of their Knowledge. If the Knowledge becomes widespread (think Excel) or if it becomes embedded in technology (think the London map and cab drivers) they become Plain Labor. So Knowledge Labor manages to capture substantial returns but lives in continuous uncertainty and need for renewal. This renewal might partly come from education, but on-the-job learning is the most important component. So workers face a catch-22 situation as they need to be part of Knowledge Labor in order to develop the skills required to be part of it. At the same time, they are forced to work long hours to keep the Knowledge advantage they have painstakingly built.
Knowledge Capital Labor. Finally, at the top of the pyramid, we have the people whose Labor infuses Capital with Knowledge. These are the hedge fund managers, private equity partners, startup CEOs, financial derivatives traders and technology executives. They can have such a disproportionate effect on the return on Capital given negative interest rates that they are rewarded for their work at a level never seen before. That is why we see the 0.1% capture an ever-increasing share of income. They are in a privileged position, but not completely immune to automation (think trading rooms substituted by algorithmic trading) or to obsolescence (think Blackberry or Nokia executives). It is a difficult club to join, with the catch-22 situation being even more extreme than with Knowledge Work. The only way to demonstrate relevant Knowledge is by managing Capital, and to manage Capital you need to demonstrate relevant Knowledge or have the Capital available. Getting and staying in the club can be gruelling. Knowledge Capital Labor has a lot of very wealthy members, that work extremely hard to keep their advantage and reputation.
This split of Labor into four different categories, two of them very difficult to enter, is stratifying society at a rapid pace. This increasing inequity is causing substantial tension in all societies, and especially in advanced democracies.
Land: from fertile fields to favourable school districts
Land continues to be valuable for some traditional uses like agriculture or resource extraction. However, a new use has emerged. Some areas, like San Francisco or London, have become Knowledge hubs. They allow people to access better Knowledge and better high-value Labor opportunities. Consequently, these areas have become extremely valuable, as the economic consequences of living in them are momentous. The Knowledge Hub effect can be found in many cities beyond San Francisco or London, even if in most it is to a lesser degree.
This has been a boon to the owners of that Land, which used to be significantly less valuable. They can extract a significant share of the economic value created by the Knowledge hubs in terms of rents. The extra income that Knowledge Labor and even Plain Labor can capture in San Francisco or London mostly goes to pay Land owners directly through rent and indirectly through rent embedded in the prices of services and products.
The death of distance might come one day and undo this effect. But for now, technology has brought an unprecedented windfall for Land owners in Knowledge hubs. It is also concentrating Labor that wants to develop the required Knowledge to play in the two upper categories in special hotspots around the world. Again, this is feeding stratification and inequity and it is creating significant tension between nation-states and regions as Knowledge hubs “hollow up” the rest of the regions.
Knowledge: implicit, explicit and mutable
Knowledge takes many different forms that we are still trying to understand. For sure a large part of it is implicit in the brains of the Knowledge Workers that use it. There is also implicit Knowledge in the processes and culture of organizations. The Apple Way or the Netflix Values or Google’s Knowledge of how modern computing works have demonstrated their capability to generate returns. This implicit Knowledge is turning the balance sheet of most companies increasingly immaterial, with property, plant and equipment representing a diminishing proportion of a company’s assets. There has been explicit Knowledge for more than a century already in the form of patents and copyright-protected content, which allow their owners to earn licensing fees. At the same time, digital technology is allowing Knowledge to be made explicit and codified at a faster rate. Any software product, like Windows, Google search or Salesforce.com, is very valuable Knowledge rendered explicit. Machine Learning models and datasets also represent codified Knowledge, which will be increasingly valuable as AI comes of age.
The other thing we know about Knowledge is that it is mutable and can be rendered obsolete or commonplace very quickly. Digital maps and navigation have substantially lowered the value of the London taxi drivers Knowledge of the city’s complex web of streets. Nokia had extremely valuable Knowledge instantiated in its phones, that was rendered obsolete by the iPhone in less than five years. Traders and Hedge Fund managers have been forced to evolve their algorithms based on human decision making to fully programmed ones that require different skills.
What is clear is that Knowledge is the key production factor in today’s world. It not only creates returns for itself but also heavily skews the returns of the traditional factors of production.
Giving this context, populism is no surprise. The majority of people don’t have access to Knowledge and are capturing a shrinking portion of the pie both from their work and savings. Trends only lead to further stratification. Knowledge Capital favours large pools of Capital, penalizing heavily the returns of small savers. Knowledge Labor and Knowledge Capital Labor are self-reinforcing making it very difficult to join them. Trying to restrict immigration or technology seems the only solution to protect the value of Plain Labor. This has brought radical options like Brexit or the Mexico Wall to mainstream acceptance. Trying to restrict capital flow to protect Plain Capital is more difficult to address because governments are dependent on low interest rates to finance their huge debt piles.
On the other side, Knowledge Capital Labor, and to a lesser extent Knowledge Labor, are very happy with the economic outcomes of the current situation. They see it as a fair consequence of their own Knowledge they work hard to develop and maintain. Becoming a millionaire or even a billionaire is now within the realm of possibility for some regardless of initial wealth. Solutions like the Universal Basic Income and free access to services are a way to secure a minimum return to Plain Labor and Attention Labor. However, they don’t address the deeper problem of meaning, and the human need to meaningful Labor.
At the same time, incredibly we are living in a time of crisis, with problems like climate change, ecological degradation, poverty and a renewed threat of conflict, maybe even nuclear conflict. Living in a world with an abundance of Capital, Labor and Land and not solving the world’s pressing problems seems incredibly stupid. We need to get our act together and mobilize our incredible resources to ensure a sustainable Earth, and maybe even reach beyond Earth. If not, we might see our stocks of Capital, Labor, Land and Knowledge quickly diminish and go back to the starting point very quickly.
Nearly 50 years ago the US Congress moved to ban tobacco ads on TV. It was finally recognized that tobacco was harmful to health. And nicotine was highly addictive, creating legions of unwilling addicts. This took a decades-long fight with the tobacco industry who tried to disprove both. The same has happened with gambling and lotteries and is lately happening with sugar. Half a century later more than 1 billion people still smoke, and there are 8 million tobacco-related deaths per year. Gambling continues to take its toll on addicts and their families. Sugar and processed foods are creating a global obesity crisis. The opioids crisis in the US is becoming a presidential campaign topic.
This happens because humans are natural-born addicts. All addictions seem to have the same physiological basis on the brain. The addictive substance triggers the release of dopamine in the nucleus accumbens which produces a pleasurable sensation. Over time, tolerance is developed and the comparatively lower release is started to be perceived as a threat by the amygdala, and thus craving starts. This physiological circuit makes addiction very difficult to overcome, even without physical withdrawal symptoms. The capability of people to overcome addiction spirals is linked to genetic sensitivity, which makes some people especially vulnerable. The environment is also critical, with exposure to the substance or behaviour making recovery very difficult.
Addicts and society pay health, economic and life fulfilment costs because of the exploitation of this addiction circuit. First, health, taking quality-adjusted years (QALYs) out of lives. For example, smoking is estimated to reduce around 3.5 QALYs by some studies. Second, economic, direct and indirect. Again, smoking has an estimated direct cost of up to $5.000 per year (varies widely per jurisdiction), health costs in the US in the high tens of billions, and estimated lower associated wages. Finally, there is a life fulfilment cost. The time we spend engaging or recuperating from the behaviour is forever lost to our loved ones and to our own development. Alcohol is a good example, with later-stage alcoholics losing their ability to function almost entirely.
Of course, drinking, smoking and gambling can be fun and enjoyable experiences. Moderate use can improve our lives without incurring heavy costs. This requires to avoid triggering the addiction circuit too strongly. The problem is that often companies are incentivized to maximize addiction. With addiction circuit triggering being directly linked to economic gain. This creates widespread mass addiction until the industry becomes regulated and incentives are aligned towards responsibility. Thus the possibility of exploiting the human addiction circuit makes regulation of industries engaging in addicting behaviour absolutely necessary.
Digital addiction: the lastest addiction circuit exploiters
Over the last 10 years, we have witnessed the blossoming of the biggest addiction yet, digital media. US adults spend more than 12 hours on average on media, more than half of which is on digital media. This means that adults spent two-thirds of their waking time engaging with media. No wonder there is less time left for engaging with our family, our friends or our community.
Digital media also has the hallmarks of addiction. We have all felt the compulsion to check one more email, scroll down one more time in the endless screen, play one more candy crush level or watch the next episode in the series. Our nucleus accumbens is in full swing, and our amygdala is ready to ping us if we try to stop. Most of us are addicted to our smartphones and our other devices and for the most part, we know it.
Digital companies also profit from addiction circuit triggering. We have many corporations that make money mostly out of our attention, Facebook, Google and Netflix being the prime examples, but many others trying to join into the party. Thus, the best brains in the world are now being funnelled to perfect and reinforce digital addiction cycles. Hundreds of thousands of engineers are working to increase advertising revenues or product engagement. They are perfecting every way in which the screen can make it slightly more difficult to turn it off. Success is keeping us some extra seconds and making us come back a little more.
This addiction is very different from others, it doesn’t have direct health costs and its economic cost is minimal. It is our life fulfilment cost that is staggering. We are sacrificing a third of our waking hours, a third of our life at the digital altar. Of course, digital media can be enjoyable and useful, there are no physical withdrawal symptoms and it is voluntary. This makes it difficult for regulators and the public to act on it. As a consequence, the smart engineers in Silicon Valley are doing their job, completely unchecked. Digital time gradually edges up crowding out other activities and making it more difficult to escape.
12 hours might seem like a lot, but the real danger lies ahead. New interfaces are being deployed: voice, Virtual Reality, Augmented Reality. They are more immersive, they are more entwined in our lives, they will be even more addictive. They might take the 12 hours closer to 18, gradually erasing any non-digital activities. If we allow digital addiction to continue unregulated we are in grave danger. The companies that make money out of engagement and their engineers won’t be able to change. They addicted to making money themselves. They think (probably correctly) “if I don’t do it, someone else will.”
Of course, the cost is not only the hours. It is also the quality of our human lives. I still have to meet someone that feels that the last four hours of digital consumption are more fulfilling than a walk in the forest, time with friends, spirituality, physical activity or creative endeavour. But the more digital expands and the rest of our lives are diminished, the harder it is to get out of the digital whirlpool.
Act quickly: lives are being squandered as we speak
We will act as a society and some of us will act as individuals. As a society, we will regulate digital addiction like we have regulated other addictions before. As always it will be difficult and it will take time. There are strong vested interests and the specifics are sufficiently different from previous addictions. While we get our act together, people lose their lives in part, or even in whole to the digital addiction. All of those human relations and endeavours that could have been but were sacrificed at the altar of engagement metrics.
For all of us, the admonition is clear. Wake up to digital addiction in our own life and that of our loved ones, and do something about it. Help wake up regulators and policymakers so digital addiction is reined in. We cannot ask the merchants of digital addiction to regulate themselves, no addiction circuit exploiter has done it before.
For those working near digital addiction, we need to wake up. Digital addiction is real and we should be helping to reduce its impact. It will require sacrifices, but good causes always do. In this time of scarce engineering talent, it is possible to make a difference and shift company policy. If you have qualms about the work you are doing, you are probably right to be concerned.
In 2012, Youtube proudly set the goal of increasing the time spent on the service from 100 million daily hours to 1 billion daily hours. In 2017, they achieved the goal ahead of schedule. Should Youtube employees be proud of their achievement? I think they should be very concerned about what they have done. Tobacco executives ingeniously pushed to sell more cigarettes to more people. Many people enjoy smoking and are able to quit at will. However, there are also many lung cancers, unwilling addicts and helpless teenagers. I am sure there are many happy stories among the billion Youtube hours. But, what are the sad stories behind those billion daily hours? What happy stories have Youtube and many others crowded out by leveraging our addiction circuit?