domingo, 28 de octubre de 2018

The Rodney Brooks (@rodneyabrooks) Rules for Predicting a Technology’s Commercial Success (via @exponentialview by @azeem)

Building electric cars and reusable rockets is fairly easy. Building a nuclear fusion reactor, flying cars, self-driving cars, or a Hyperloop system is very hard. What makes the difference? 
The answer, in a word, is experience. The difference between the possible and the practical can only be discovered by trying things out. Therefore, even though the physics suggests that a thing will work, if it has not even been demonstrated in the lab you can consider that thing to be a long way off. If it has been demonstrated in prototypes only, then it is still distant. If versions have been deployed at scale, and most of the necessary refinements are of an evolutionary character, then perhaps it may become available fairly soon. Even then, if no one wants to use the thing, it will languish in the warehouse, no matter how much enthusiasm there is among the technologists who developed it. 
Here I present a short list of technology projects that are now under way or at least under serious discussion. In each case I’ll point out features that tend to make a technology easy or hard to bring to market.
  • Not Much Needs to Change
  • Haven’t Been There, Haven’t Done That
  • Obstacles are closer than they appear
  • No Component Is Too Hard, but All Together They’re a Bear
  • Sometimes, the Possible Just Takes a Little Longer 
In pointing out the differences that make one technology harder than another, I am not preaching technological defeatism. I’m only suggesting that we properly gauge the difficulty of whatever we are told could be the next big thing. If the idea builds on practical experience, then guarded optimism is in order. If not, then not. Hope is a scarce thing; we shouldn’t squander it.

lunes, 1 de octubre de 2018

Resumen–comentario de "Tecnología vs Humanidad" (por @respla para @sintetia)

UPDATED: Nuevos comentarios de @respla ( https://www.sintetia.com/tecnologia-vs-humanidad-de-gerd-leonhard-con-o-contra-ii/ ) autor de la entrada que resumo a cotinuación.

2ª y última actualización para aportar el cierre de las consideraciones de @respla sobre el libro "Tecnología vs Humanidad" de Gerd Leonhard.

**
Así queda la reseña del libro en las ideas generales –abreviadas en esta entrada–, y enlaces a la segunda y  tercera parte, destinadas a comentar los aspectos más debatibles la 2ª mientras que la 3ª revisa alguna de las ideas propuestas para "minimizar los riesgos del momento de explosión tecnológica en el que vivimos."


Tecnología vs Humanidad de Gerd Leonhard: ¿con o contra? – @Sintetia otra interesante entrada que no os debéis perder en su totalidad…

Extractos a continuación (con énfasis míos):
…nos encontramos en un punto de inflexión para la humanidad. Un momento crítico creado por los cambios tecnológicos acelerados que vivimos, y sobre todo en lo relativo a lo digital.

…define tres características comunes que tienen este tipo de tecnologías transformadoras:
1.- Exponencial: Los avances tecnológicos siguen curvas parecidas a las de la Ley de Moore. Esto genera un reto cognitivo enorme a los seres humanos ya que nosotros seguimos formas lineales de aprendizaje.
2.- Combinatorio: Las tecnologías se están combinando y convergiendo entre sí para conseguir avances aún más rápidos.
3.- Recurrente: Hay tecnologías que aprenden por sí mismas y que cada vez necesitan menos de humanos para su mejora. 
A su vez, Leonhard describe los 10 megacambios tecnológicos que estamos viviendo en la actualidad, cambios fruto de la convergencia entre tecnologías exponenciales que se están desarrollando simultáneamente: 
  1. Digitalización: Todo lo que pueda ser digitalizado, será digitalizado  
  2. Movilización: Sin cables, móvil y siempre conectado. Esto nos lleva también a que todo se graba.
  3. Pantallización: Revolución de los interfaces.
  4. Desintermediación: Capitalismo de plataformas.
  5. Transformación: La verdad detrás del término ya vacuo de “transformación digital”.
  6. Inteligización: Las cosas se están volviendo inteligentes.
  7. Automatización: Cuando las cosas son inteligentes, luego se automatiza.
  8. Virtualización: Crear una versión digital de todas las cosas.
  9. Anticipación: Las máquinas nos ayudarán a predecir como nunca antes
  10. Robotización: La materialización de todo esto

Leonhard considera que hay cinco etapas en el proceso por el cual vamos cediendo protagonismo a las máquinas, primero en las acciones y luego en las decisiones. Estas cinco etapas son: 

  1. Automatización: Exponencial e inevitable. Pero, ¿debería esta eficiencia realmente prevalecer sobre la humanidad? ¿Deberíamos automatizar las cosas por el simple hecho de que podamos hacerlo?   
  2. Asentimiento: La aceptación de sistemas que nos sustituyen en determinadas acciones porque nos lo hacen fácil y cómodo. Por ejemplo, utilizar sistemas que escriben mensajes por nosotros. 
  3. Abdicación: Renunciamos a hacer cosas que eran de nuestra responsabilidad y las delegamos en máquinas. Confiamos ciegamente en las recomendaciones de las máquinas. 
  4. Agravio: Discriminamos a los seres humanos frente a las máquinas o frente a seres humanos “aumentados”. 
  5. Abominación: El momento de la despersonalización total en el que ya no vemos otras personas sino que vemos números, recomendaciones y evaluaciones dadas por máquinas. 

Por supuesto que el autor se sitúa en el campo humanista y sus propuestas en este ámbito son la parte central del libro. De esta manera, Leonhard advierte de sustituir lo que nos hace humanos, los “androritmos”, por algoritmos lo cual pone en peligro nuestra humanidad. 
En esta línea sugiere reforzar el CORE (creatividad/compasión, originalidad, reciprocidad/responsabilidad y empatía) frente al empuje de las STEM (ciencia, tecnología, ingeniería y matemáticas). Estos androritmos entroncan también con la ética, algo que en principio no sabemos si las máquinas podrán desarrollar algún día. 

Creo que la reflexión más importante del libro se sitúa alrededor de la felicidad. Aporta una afirmación muy interesante: el objetivo principal del progreso tecnológico debería ser la búsqueda de la máxima felicidad humana. Este gran objetivo trae consigo la gran pregunta: qué es la felicidad. 
Se suele hablar de dos tipos de felicidad, la hedonista y la eu̯dai̯monía. La primera es la del ahora y la de los placeres. La segunda tiene más que ver con la prosperidad. 

El psicólogo Martin Seligman utiliza el modelo PERMA para hablar de la verdadera felicidad que no viene sólo de placeres externos y momentáneos: 

  • Pleasure (placer): comida sabrosa, baños calientes. 
  • Engagement (compromiso): Participar en actividades desafiantes. 
  • Relationships (relaciones): los vínculos sociales han mostrado ser un indicador extremadamente confiable de la felicidad. 
  • Meaning (sentido): una búsqueda percibida de pertenencia a algo más grande que nosotros mismos. 
  • Accomplishments (logros): haber alcanzado metas tangibles.
Corremos un riesgo importante de que la tecnología evolucione tanto como para simular toda estas fuentes de felicidad hasta tal punto que no podamos distinguirlas de la realidad.  

El autor acaba el libro con algunas predicciones que dibujan un mundo distópico y que parecen realizables viendo el estado actual de la tecnología y su posible evolución, así como con la recomendación de la formación de un Consejo Global para la Ética Digital (CGED) con la tarea de definir cuáles serían las reglas base y los valores más primordiales y universales que una sociedad tan radicalmente diferente y digitalizada debería tener. 

Para apoyar este CGED y comenzar un debate sobre el tema de la ética digital, el autor habla de un futuro manifiesto que impulsara su creación, para el que propone cinco derechos fundamentales: 

  1. El derecho a seguir siendo naturales, esto es, biológicos.
  2. El derecho a ser ineficientes si esto define, o cuando defina, nuestra humanidad básica. 
  3. El derecho a desconectarnos. 
  4. El derecho a ser anónimos. 
  5. El derecho a emplear o involucrar a personas en lugar de máquinas.

sábado, 29 de septiembre de 2018

Compassion & Technology and the Life of the Buddha (from @singularityunl)

“Our earth will eventually disappear, our sun will disappear, even our galaxy will ultimately disappear, so it’s unrealistic to think we will avoid death.”
https://singularityuthenetherlands.org/news/compassion-technology-and-the-life-of-the-buddha-at-the-nieuwe-kerk/


asked about how technology and compassion could be of help to other people around the world, he answered: 
“Machines are very important, but they are controlled by human beings. We human beings are not only physical entities, we also have minds. When we are motivated by positive emotions our physical actions will be constructive. Modern psychology knows about sensory consciousnesses, but doesn’t distinguish them clearly from mental consciousness, which involves emotions like anger. I’m very appreciative of the comfort and relief that technology can provide, but I’d like to see its effects implemented in less developed countries where there is still great suffering.” 
“These machines are material devices,” His Holiness observed, “but we also have to think about consciousness. Our waking consciousness depends on our brain and sensory organs and is relatively coarse. When we dream the senses are at rest. In deep sleep, consciousness is subtler, as it is when we faint and so forth, but the subtlest, deepest consciousness manifests at the time of death.

His Holiness explained that psychologist Richie Davidson of University of Wisconsin–Madison has undertaken a project to investigate what is going on. He pointed out that while technology can improve eye and ear consciousness, it has little effect on the subtler level of mental consciousness that nevertheless can be extended infinitely. Inner values involve the mind and ancient India was rich in understanding the mind’s workings as a result of the practices for cultivating a calmly abiding mind (shamatha) and analytical insight (vipashyana). The Buddha’s attainment was a product of such practices. 
 
The challenging question raised was, “Would you like to live to be 1000 years old?” 
His Holiness retorted that it’s necessary to be realistic and the question represented unrealistic thinking. He observed that Indian Sadhus and others have tried to achieve such a goal through yoga and breath control, but none have lived more than 200 years. 
 
Selma Boulmalf asked His Holiness if sickness had any meaningful role in life
He told her he thought that facing pain and difficulty reminds believers of God and their religious path. 
…He explained three levels of knowledge: basic understanding gained by hearing or reading, conviction that derives from critical thinking and experience arising from deeper acquaintance in meditation. 
His Holiness told her that even animals love life and move to defend it. “We all naturally love life and death brings an end to it. We tend to fear death because it is a mystery, but through training we can develop confidence in the next life.”   
We need to focus on a sense of the oneness of humanity and maintaining religious harmony, which India vividly exemplifies. If religious harmony can flourish there, why not elsewhere?”
 
Question to His Holiness was about whether there has been a female Dalai Lama and if not, could there be one in the future? 
His Holiness replied that he had been asked this repeatedly over the years and has answered that if a female body would be more useful, why not? He qualified this by adding that whether or not there will continue to be a Dalai Lama in the future is something Tibetans, Mongolians and people of the Himalayan Region will decide.

During a meeting with members of the media immediately afterwards His Holiness remarked that technology can clearly play a significant role in alleviating physical distress, but that peace of mind and the role of moral principles cannot be overlooked. He commented that existing education sets material goals, resulting in aspirations for a more materialistic way of life with little attention to inner values.
 
He expressed disapproval of the use of technology for oppressive surveillance, but noted that the problem lies with the motivation of perpetrators and the way it is used rather than the technology by itself. He repeated that moral principles lay the basis for individuals, families and communities to live a happy life.

domingo, 23 de septiembre de 2018

A manifesto for renewing liberalism (by @TheEconomist)

via @jorgecortell A manifesto for renewing liberalism – The Economist

reinvention is always a good idea in my opinion
Liberalism made the modern world, but the modern world is turning against it. Europe and America are in the throes of a popular rebellion against liberal elites, who are seen as self-serving and unable, or unwilling, to solve the problems of ordinary people. Elsewhere a 25-year shift towards freedom and open markets has gone into reverse, even as China, soon to be the world’s largest economy, shows that dictatorships can thrive. 

Our founders would be astonished at how life today compares with the poverty and the misery of the 1840s. Global life expectancy in the past 175 years has risen from a little under 30 years to over 70. The share of people living below the threshold of extreme poverty has fallen from about 80% to 8% and the absolute number has halved, even as the total living above it has increased from about 100m to over 6.5bn. And literacy rates are up more than fivefold, to over 80%. Civil rights and the rule of law are incomparably more robust than they were only a few decades ago. In many countries individuals are now free to choose how to live — and with whom. 
Liberalism emerged in the late 18th century as a response to the turmoil stirred up by independence in America, revolution in France and the transformation of industry and commerce. Revolutionaries insist that, to build a better world, you first have to smash the one in front of you. By contrast, conservatives are suspicious of all revolutionary pretensions to universal truth. They seek to preserve what is best in society by managing change, usually under a ruling class or an authoritarian leader who “knows best”. 
An engine of change 
True liberals contend that societies can change gradually for the better and from the bottom up. They differ from revolutionaries because they reject the idea that individuals should be coerced into accepting someone else’s beliefs. They differ from conservatives because they assert that aristocracy and hierarchy, indeed all concentrations of power, tend to become sources of oppression. 
Liberalism thus began as a restless, agitating world view. Yet over the past few decades liberals have become too comfortable with power. As a result, they have lost their hunger for reform. The ruling liberal elite tell themselves that they preside over a healthy meritocracy and that they have earned their privileges. The reality is not so clear-cut. 
 
In all sorts of ways, the liberal meritocracy is closed and self-sustaining. A recent study found that, in 1999–2013, America’s most prestigious universities admitted more students from the top 1% of households by income than from the bottom 50%. In 1980–2015 university fees in America rose 17 times as fast as median incomes. The 50 biggest urban areas contain 7% of the world’s people and produce 40% of its output. 
 
It is the moment for a liberal reinvention. Liberals need to spend less time dismissing their critics as fools and bigots and more fixing what is wrong. The true spirit of liberalism is not self-preserving, but radical and disruptive.
 
They must rediscover their belief in individual dignity and self-reliance — by curbing their own privileges. They must stop sneering at nationalism, but claim it for themselves and fill it with their own brand of inclusive civic pride. Rather than lodging power in centralised ministries and unaccountable technocracies, they should devolve it to regions and municipalities. Instead of treating geopolitics as a zero-sum struggle between the great powers,…
 
The best liberals have always been pragmatic and adaptable. …
Liberals should approach today’s challenges with equal vigour. If they prevail, it will be because their ideas are unmatched for their ability to spread freedom and prosperity. Liberals should embrace criticism and welcome debate as a source of the new thinking that will rekindle their movement. They should be bold and impatient for reform. Young people, especially, have a world to claim.

It’s learning. Just not as we know it. (by @accenture)

Skills Gap in the Future Workforce | Accenture
“We still talk about a knowledge economy, but the reality is that the world is moving beyond it. What we have now is an innovation economy. Knowledge has been commoditized. There is no longer a competitive advantage in simply knowing more than other people, because Google knows everything. What the world cares about is not how much you know, but what you can do with it.” 
~Tony Wagner, Senior Research Fellow, Learning Policy Institute. 
Full pdf report link.











domingo, 9 de septiembre de 2018

China Is Quickly Becoming an AI Superpower (by @singularityhub)

Propelled by an abundance of government funds, smart infrastructure overhauls, leading AI research, and some of the world’s most driven entrepreneurs, China’s AI ecosystem is unstoppable.


As discussed by Kai-Fu Lee in his soon-to-be-released book AI Superpowers, four main drivers are tipping the balance in China’s favor… 
1. Abundant dataPerhaps China’s biggest advantage is the sheer quantity of its data. Tencent’s WeChat platform alone has over one billion monthly active users. That’s more than the entire population of Europe. 
Take mobile payments spending: China outstrips the US by a ratio of 50 to 1.
…While the US saw $112 billion worth of mobile payments in 2016, Chinese mobile payments exceeded $9 trillion in the same year.   
 
2. Hungry entrepreneurs empowered by new toolsFormer founder-director of Google Brain Andrew Ng noted the hunger raving among Chinese entrepreneurs: “The velocity of work is much faster in China than in most of Silicon Valley. When you spot a business opportunity in China, the window of time you have to respond is very short.” 
But as China’s AI expertise has exploded, and startups have learned to tailor American copycat products to a Chinese audience, these entrepreneurs are finally shrugging off their former ‘copycat’ reputation, building businesses with no analogs in the West.
 
3. Growing AI expertiseIt is important to note that China is still new to the game. When deep learning got its big break in 2012—when a neural network decimated the competition in an international computer vision contest—China had barely woken up to the AI revolution. 
But in a few short years, China’s AI community has caught up fast. While the world’s most elite AI researchers still largely cluster in the US, favoring companies like Google, Chinese tech giants are quickly closing the gap. 
Already in academia, Chinese AI researchers stand shoulder-to-shoulder with their American contemporaries. At AAAI’s 2017 conference, an equal number of accepted papers came from US- and China-based researchers. 

4. Mass government funding and supportThe day DeepMind’s AlphaGo beat top-ranking Chinese Go player Ke Jie has gone down in history as China’s “Sputnik Moment.” 
Within two months of the AI’s victory, China’s government issued its plan to make China the global center of AI innovation, aiming for a 1 trillion RMB (about $150 billion USD) AI industry by 2030.

lunes, 27 de agosto de 2018

The Startup vs Enterprise QUEST (by @saranormous via @greylockvc)

 Startups Serving The Enterprise: – Greylock Perspectives



Building strong partnerships and capabilities means that getting out of the marketing swamp, through the winds of cost and risk, across the enterprise feasibility gap, through the desert of procurement and over the ocean of early execution — will all be more tenable the second time around, and the rewards even richer on both sides.

Why are large enterprises so interested in startup tech? It’s a matter of survival. Every company is undergoing a digital transformation. Farsighted executives see the pace of change in business accelerating. These executives know that the companies who more rapidly adopt advancing technology will run their companies better, faster, cheaper, smarter. Technology is a weapon used to defend against competitive threats, and achieve and preserve market dominance.

Similarly, for an enterprise technology startup to survive and thrive, they must understand how to effectively work with large companies. Within large enterprises are most of the employees, data, workflows, industry and institutional knowledge, assets, customer relationships, intellectual property, and budgets in the world.


1. Recruiting Partners in the Swamp of Marketing Fog
Before enterprise executives and startup founders are ready to set sail together, they need to identify the right partners.

  • Shine a Bright Light: A warm introduction can be vital.
  • Paint a Clear (and Easy) Path Out: Startups need to clearly explain the problem they are solving, their value proposition, and their differentiation.
  • Seek the Right Stakeholders, at the Right Time: And those other technologists and leaders are structurally more aggressive in technology adoption than the CIO.

2. Maintaining Faith through the Galewinds of Cost and Risk
One enterprise tech leader said that talking to his team about bringing in a new technology inevitably triggers an immune defense reaction: New vendors need to understand how customers are measuring return and cost. 

Even once your team has cleared a path out of the swamp, it can feel like you’re fighting against a galewind. There’s a lot of natural resistance to bringing in new vendors, because a new offering needs to be valuable enough to overcome inherent cost and risks of working with a startup.

Startups should be empathetic to this risk aversion and understand that, on the customer side, someone’s career is often on the line.

Enterprise customers told us they think also about the “hidden costs” of vendor management, user training and adoption, integration, implementation and administration, and the risk of the startup dying or getting acquired.
Because of these many “hidden” costs, smart technology buyers are projecting out the landscape of vendors, and looking for startups that not only offer tactical benefits but have a chance to endure as longer-term partners — disrupting an existing category or creating an important new one.
Advantages for disruptor companies include innovating on the experience of purchasing and using the technology, and the total cost of ownership.

To de-risk their decisions, enterprise tech leaders want to work with startups that have raised capital from top-tier investors, because it’s one sign they’ll go the distance.


3. Bridging the Gap of the Four S’s: Scale, Security, Spend & Supportability
Value may outweigh the costs and risk, but will the product work in their environment?

  • Scale: Can the startup support the scale of the customer’s user base or infrastructure? Increasingly, we see customers want to validate that scale rather than taking it on faith. … This includes ease of use, rollout plan, reporting, integrations into existing technology, administration workflows, SLA’s. Customers are also evaluating who is going to help them deploy — for example, the existence and quality of the startup’s sales engineering or implementation team, if needed.
  • Security: Startups told us this is #1 on everyone’s list. The need is often driven by regulation such as GDPR, or internal requirements for sophisticated access control, and the key thing here is to have a clear approach to customer data.
  • Spend: Pricing models that are appropriate for the first thirty developers or first hundred users might not work for broad deployment. Startups must offer pricing models that are feasible at scale, aligning with the value they create for their customers.
  • Supportability: Is the startup prepared to offer the kind of support (often 24/7) that enterprise customers need, at scale? Can they handle the reality of legacy technology that large companies are often saddled with, and make their customer successful?

4. Avoiding the Quicksands of Customization

The quicksands of customization are an especially tricky neighborhood.
…getting sucked into customization can mean company death, or at least, derailment.

Just as customers will choose to work with a particular startup based on ability to scale, durability, and other factors, startups should also choose their early customers carefully, balancing customer requirements against strategic priorities and limited company resources. Being too accommodating or diffuse in strategy can be a recipe for mediocrity in multiple categories.
Disciplined customer segmentation is key

Giving potential customers realistic visibility into your short and medium term roadmap is also a pattern for success. Beyond choosing early customers carefully, startups should also take a pragmatic view of what feature requests to field, and when.


5. Surviving the Desert of Procurement & Approvals
The procurement process can be a bear — you feel like you’re so close to the finish line, but it’s a mirage. You can get stuck in limbo.
Startups need to have realistic expectations about speed, and plan ahead for sales cycles so they don’t run out of resources before they show progress.
Enterprises, on the other hand, need to create pathways for the business to push through important innovations fast.

To accelerate their sprint through the procurement and legal desert, startups should find internal champions, arm those buyers with the right business case and other support, and arrive prepared with mature contracts.
There are also different purchasing processes for different scales of spend. Building up engagement with a large enterprise partner through a land-and-expand model also changes a startup’s initial experience in procurement.


6. Crossing the Ocean of Early Execution
Finally, quest-goers need to build a strong ship and chart a clear course to cross the ocean of early execution.

  • First, this means structuring PoCs and initial engagements to be short, with repeatable onboarding flow, clear success criteria and commitment from partners to that timeline.
  • Second, technology leaders also cautioned against “poisoning the well” — damaging relationships and reputation by not delivering on promises.
  • Third, enterprises need startups to consciously involve the necessary stakeholders to operationalize technology, even in planning and deployment, support their change management, and follow up with discipline.

7. The Golden Fields of Innovation

In summary, the quest to reach the golden fields of innovation — that is, to successfully deploy new capabilities and technologies into the enterprise — is a journey that requires strategy, careful planning and consistent execution.
Once that early execution is successful, this is not a one-time quest. It’s an ongoing journey with each new partner, and each new use case and product line. An early success lays the groundwork for a strong customer reference that will help generate new business — in today’s age of connectedness and transparency, a startup’s best salespeople are its happy customers.

domingo, 26 de agosto de 2018

With Greed and Cynicism, Big Tech is Fueling Inequalities in America (by @filloux)

With Greed and Cynicism, Big Tech is Fueling Inequalities in America – Monday Note, Frederic Filloux

…In the end, local taxpayers will subsidize Amazon shareholders…

Hi-tech firms are prominent among recent tax-break “megadeals” awarded by cities and states. Tesla’s battery factory ($1.3bn from Nevada), Foxconn’s display-screen plant in Wisconsin ($4.8bn) and Apple’s data centre in Iowa ($214m) are typical. The Apple centre, a cloud computing facility, will have only 50 permanent jobs, so the cost per job exceeds $4.2m. The Foxconn deal, even by the state’s own official estimate, won’t break even for taxpayers for 25 years — an extremely risky time horizon given the likelihood of new technologies leapfrogging the company’s product much sooner. The Tesla deal was 14 times costlier than anything Nevada had done before.

5 online-luxury-fashion trends (by @forbes via @wealthx)

  What Farfetch's IPO Filing Says About The $300 Billion Luxury Fashion Industry – Wealth-X



1. By 2025, luxury fashion e-commerce will have a quarter of the industry's market share. 

Industrywide, online’s share of the personal luxury goods market is expected to rise to 25% by 2025 from about 9% last year as the entire market pie will see 45% growth to $446 billion over the same period, from $307 billion last year, Farfetch said in the filing, citing a Bain & Co. study.


2. Yes, credit or (blame) Millennials, and soon Gen Z.

By 2025, the two groups combined, mostly led by Millennials, will represent 45% of total luxury spending. That will beat the 40% share expected to be held by Gen Xers


3. Democratization of luxury fashion?

…online sales have leveled the playing field for young brands seeking access to consumers, marketplaces are giving smaller luxury brands and boutiques a bigger opportunity to reach a global consumer base.ç


4. Data science is hot.

“We are a technology company at our core. We operate at the intersection of luxury fashion, online commerce and technology." (Farfetch)
What kind of data does it have? real-time inventory data, global behavioral and transactional data and pricing data for over 335,000 SKUs (unique units) from more than 3,200 different brands…


5. Growth at the cost of profit may be the new norm.

Amazon’s years-long model of investing behind growth before it finally turned a profit looks to be increasingly popular among companies eyeing online growth and seems to be accepted by investors.