ºÎ»ê½Ãû µµ¼­¿ä¾à
   ±Û·Î¹ú Æ®·»µå³»¼­Àç´ã±â 

åǥÁö






  • Why 2021 is Nothing Like 2000 or 2008

    Human beings have a hard-wired ¡°availability-bias.¡± Reflexively, we see every major challenge through the lens of the previous challenges we¡¯ve personally faced. For Xers and Millennials, there¡¯s an inevitable fear that the next crisis will be a repeat of the 2008-to-2009 sub-prime crisis. For Boomers, the subprime crisis is familiar, but clearly not as ¡°relevant¡± as the Dot-Com crash of 2000.

    That¡¯s really dangerous because few of today¡¯s decision-makers have first-hand experience of the Golden Age of the Mass Production Techno-Economic Revolution which unfolded after World War II. That period of unprecedented economic growth and rising prosperity is almost forgotten. And yet, it is far more relevant to today¡¯s situation than the SubPrime Crisis or even the Dot-Com Crash.
    To understand why, it¡¯s important to remember that until the end of the 18th century, affluence was based on agriculture and it had remained at persistently low levels for thousands of years. Then, beginning in around 1770, five successive waves of revolutionary technological change, raised standards of living to an extent that was previously unimaginable. Each of these revolutions was based on a transformational technological advance such as ¡°the steam engine,¡± ¡°the railroad,¡± or ¡°the assembly line.¡±
     
    And each of these five techno-economic revolutions consisted of three phases called ¡°the installation phase,¡± ¡°the transition phase¡± (or turning-point) and ¡°the deployment phase.¡± Each installation phase began with a ¡°technological big bang¡± in which a new general-purpose core technology suddenly appeared on the scene; for instance, the installation phase of the fifth techno-economic revolution began with the introduction of the Intel 4004 microprocessor in 1972.

    As applications of the revolutionary technology began to proliferate, each installation phase inevitably became a time of ¡°speculative frenzy.¡± In the case of the Fifth Techno-Economic Revolution, this frenzy ended in the Dot-Com crash.

    Why does the installation phase of every techno-economic revolution inevitably end in a speculative crash? Because the reality of the technologies, business models, and underlying institutions which are currently available falls short of being able to cost-effectively deliver on the promises made to investors. In short, the new techno-economic paradigm will not be ready to transform the economy for a long time. A great deal of fundamental innovation is still required. And that often means that a lot of painful creative-destruction is required; for instance, exponentially increasing the bandwidth of 90s-era fiber networks to enable today¡¯s gigabit streaming platforms. In 2000, search engines, smart phones, and speech recognition, all had to evolve from expensive concepts to high-performance, mass market solutions. They did so during the transition phase. And as a result, both the supply and demand sides of the business equation are finally poised to enable radical change and explosive growth. That¡¯s the catalyst for the Golden Age.

    In previous issues of Trends , we¡¯ve examined nine complementary technologies which finally have the ¡°critical mass¡± required for explosive commercial take-off between 2020 and 2025:

    1. Deep Learning & Artificial Intelligence;
    2. New-Era Computing Platforms (including GPUs as well as ARM & RISC-V processors);
    3. Virtual & Augmented Reality;
    4. Business Automation (in Factories & Services);
    5. Autonomous cars & trucks;
    6. Commercial Unmanned Aerial Vehicles;
    7. Orbital Aerospace Solutions;
    8. Additive Manufacturing; and
    9. 3rd Generation Genomics solutions including LRS, Cancer Screening & Gene Therapy.

    And it¡¯s not just a matter of these technologies being ready to deliver benefits. Consumers and businesses are flush with cash and eager to invest in and consume these solutions. Here again, today¡¯s environment resembles the post-war Golden Age.

    The resulting opportunities are real. And retail investors can identify and profit from them, NOW! The fact that many market commentators are comparing the current stock market environment with the 2000 Dot-Com bubble without appreciating the profound differences actually creates an enormous opportunity for those who aren¡¯t fooled.

    Why? Because seeing opportunities to which other are blind enables you to ¡°be greedy when others are fearful.¡± And, according to Warren Buffett that¡¯s the key to long-term superior results.

    Given this trend, we offer the following forecasts for your consideration.

    First, by 2037, deep learning & artificial intelligence will create $30 trillion in cumulative market capitalization. As the Trends editors have reiterated over the past decade, this transformative general-purpose technology will finally unleash the game-changing potential of the Fifth Techno-Economic Revolution and overcome the demographic drag created by an aging workforce. As of year-end 2020, deep learning was estimated to have already added a total of $2 trillion in market capitalization, worldwide. However, that¡¯s just the beginning! ARK Invest estimates that this will grow at a compound rate of 17% a year through 2037; at that point, machine learning and AI will represent roughly $30 trillion of global market capitalization or 15% of all stock market value. That means deep learning and AI will have created 50% more cumulative value than the Internet, in far less time. Why is this happening NOW? Because of a convergence of inexpensive special-purpose processors, advanced software methods, and mountains of accumulated data that can be used to train applications. And make no mistake, it¡¯s companies with access to proprietary data which will reap the biggest rewards, rather than the companies which invent the AI-tools.

    Second, as the cloud expands over the coming decade and hosts much of this AI capability, the biggest hardware opportunity lies in so-called ¡°processing accelerators¡± such as Graphics Processing Units, Tensor Processing Units, and Field-Programmable Gate Arrays, which perform the most demanding functions cheaply. By 2030 annual sales of ¡°server accelerators¡± could reach $41 billion a year. Over the next five years, revenues from Graphics Processing Units will grow faster than any other category of cloud hardware, largely because GPUs play such a crucial role in AI, big data analytics and cloud gaming. NVIDEA seems especially well-positioned to benefit in this and other competitive spaces. While CPUs will continue to grow in power and number, falling prices will mean that CPU revenues will actually shrink while ARM & RISC-V CPUs will largely replace Intel¡¯s X-86 processors.

    Third, over the next five years, the global market for Virtual & Aug mented Reality will grow 59% annually from just $3.5 billion to roughly $28 billion a year. Nothing illustrates the difference between 1990s and 2020s technologies better than the status of VR: When the Trends editors first saw a demonstration of virtual reality in 1992, we proclaimed that it was ¡°20-to-30 years away from the commercial mainstream¡± and we noted that ¡°gaming was a just a niche market.¡± But over the next five years, AR & VR are estimated to grow at a compounded annual rate of 59% and worldwide gaming revenues are expected to approach $400 billion annually.

    Fourth, business automation alone could add $1.5 trillion to 2025 U.S. GDP while increasing economic growth by 1% a year. While widespread uncertainty about global growth interrupted the accelerating demand for factory robots in 2020, this is now behind us. Low interest rates, booming consumer demand and falling robot prices make automation increasingly attractive. Furthermore, recent strides in machine learning applied to training robots is finally making factory automation cost-effective for a whole category of smaller companies. Over the next five years, automation will directly raise manufacturing productivity by reducing labor¡¯s share of manufacturing costs. This will result in higher wages, lower prices, higher margins and higher capital investments.

    Fifth, fleets of autonomous cars & trucks will transform the economics of ground transportation around the world in the decade ahead. For over 100 years, user-owned automobiles have dominated transportation in the developed world. Now suddenly, self-driving technology is poised to provide a safer, cheaper, more reliable alternative. While self-driving cars were imagined at least as far back as the 1950s, it wasn¡¯t until the emergence of Internet-based Ride-Hailing that cost-savings and consumer-acceptance began to converge. Today¡¯s American ride-hailing model has mass appeal at $1.85/mile, but studies show that autonomous ride-hailing could deliver an equally profitable ride for $1.00/mile. And while there are still some regulatory and technical barriers to overcome, it seems clear that one or more of the three technical alternatives embraced by Tesla, Baidu, Alphabet¡¯s Waymo and General Motors will be ready for commercial deployment as early as 2022. And notably, much of Tesla¡¯s current valuation rests on its promise to rapidly ramp up deployment of autonomous ride-hailing ahead of competitors.

    Sixth, during the 2020s, commercial drones carrying passengers and packages will finally overcome regulatory barriers that have kept them grounded. Slowly and steadily unmanned package delivery drones have overcome an array of technical hurdles. Consumer attitudes, regulatory priorities, technological infrastructure and vehicle performance are all rapidly converging to enable limited commercialization by 2025 and widespread market penetration by 2030. Companies like Amazon, FedEx and UPS are all ready to pounce on the package delivery space. Meanwhile, Lilium and Joby (the successor to Uber Elevate) are likely to lead the way in offering airborne ride-hailing services. Across the spectrum, drone-based alternatives, have a clear advantage over conventional taxi and package delivery solutions in terms of speed and cost. By 2030, ARK¡¯s estimates indicate that drone delivery platforms will generate $275 billion a year in service revenues, $50 billion in hardware sales and $12 billion in mapping revenues.

    Seventh, Orbital Aerospace Solutions will become a major business opportunity as launch costs drop by a factor of 200 in the decade ahead. Payload launch costs to Low-Earth Orbit (or LEO) which were $65,400/kg with the Space Shuttle are currently at $2,600/kg for the SpaceX Falcon 9. But now, SpaceX is talking about $10/kg with its new Starship. And because Starship is designed to be refuellable in orbit, space planners will be able to launch 150 tons to LEO, refuel while orbiting Earth, and then fly the same payload the rest of the way to the moons of Jupiter. That¡¯s exciting. But the real impact may be more pedestrian. Meta-analyses have found that trade volume (on Earth) has a roughly inverse-linear relationship to transport costs. If that relationship holds true for space, a 200-times cost reduction in travel between Earth and LEO should increase ¡°trade¡± between Earth and LEO by 200-times. Commerce between the Earth and the moon, or between the Earth and Mars, starting from a base close to zero, would be stimulated even more. We will have to wait to see exactly what this portends. Notably, SpaceX is already testing prototypes of the Starship platform.

    Eighth, over the decade ahead ¡°additive manufacturing¡± will become a $600 billion a year opportunity as it begins to increasingly address the manufacture of end-use parts. Initially, additive manufacturing focused on prototyping, where speed was not crucial. But going forward, the largest market segments will be end-use parts for aerospace as well as molds & tools for auto parts and machinery. — And that¡¯s just the beginning: As we discuss this month in trend #4, today¡¯s pioneering research in bio-printing will leverage other advances in additive manufacturing to produce the first commercial organs for transplantation to humans as soon as the end of the decade. And,

    Ninth, the 2020s will see so-called ¡°3rd Generation Genomics¡± revolutionize the economics of health care by exploiting LRS sequencing, socalled liquid-biopsies and improved gene therapies. First, reading a genome using Long-Read Sequencing (or LRS) can provide dramatically better information than the ¡°shot-gun approach¡± currently used; by 2023 the cost of LRS sequencing is expected to fall from $3500 to under $1,000 per genome, opening-up mass-market demand. Second, a single multi-cancer screening blood test (aka ¡°Liquid Biopsy¡±) can detect dozens of early-stage cancers; by 2025 the prototype test that cost $30,000 in 2015 is expected to be commercially available for just $250 per patient, opening-up mass-market demand for another transformational solution. And third, by shifting from personalized gene therapy using one¡¯s own cells to (so-called Allogeneic) gene therapies based on ¡°off-the shelf cells,¡± costs will be reduced by 90% and incremental revenues will grow by $250 billion a year. What¡¯s the bottom line? This wave of innovation will enable a dramatically healthier population with many more productive years to produce, consume and enjoy themselves.

    Resource List
    1. Edward Elgar Publishers. April 26, 2003. Carlota Perez. Technological Revolutions and Financial Capital: The Dynamics of Bubbles and Golden Ages.

    2. Trends. January 15, 2021. Trends Editors. Putting the Great Stagnation Behind Us.

    3. Trends. January 15, 2021. Trends Editors. The Tumultuous Birth of Our New Golden Age.

    4. Trends. June 15, 2012. Trends Editors. Understanding the Great Inflection Point Ahead.

    5. Trends. August 15, 2015. Trends Editors. America¡¯s 4% Growth Challenge.

    6. Ark Invest. January 26, 2021. ARK Investment Management LLC. Big Ideas Report 2021.