Key Takeaways
- 20th-century principles of mass production and economies of scale are ceding to mass personalization and rentable scale.
- Rentable scale is allowing small, nimble startups to take on entrenched incumbents across every sector of the economy.
- Data and algorithms are making it possible to create and iterate on services that are more personalized, higher quality, and fundamentally better than the mass market products and solutions available to us today.
- AI will drive innovation but it has to be used responsibly. The potential for significant negative impact of the misuse or misunderstanding of algorithms is only just coming to light.
- We will need to create a significant public/private partnership to make the most of AI-driven innovation while avoiding the potential pitfalls.
The book Unscaled by Hemant Taneja explores how startup companies can create capabilities similar or stronger than large companies by unscaling. Such companies compete by renting space and functionality in the cloud, which makes them cheaper and more flexible. They are able to innovate and create better products by using data and exploiting the possibilities that sophisticated AI is increasingly offering.
InfoQ interviewed Taneja about what "unscaling" is, the things that startup companies can rent and how that gives them an advantage, the breakthrough of Artificial Intelligence (AI) and how companies apply AI to make products which better meet customer demands, what AI will bring us in the future, and the benefits that unscaling can bring.
InfoQ: What made you decide to write this book?
Hemant Taneja: When I first moved to Silicon Valley six years ago, I began meeting entrepreneurs who were interested in tackling society’s biggest challenges. Beyond just creating software to find efficiencies for businesses, they were taking on health care, finance, education. What I found immediately fascinating was how these tiny startups were able to spin up so quickly to create companies with the capabilities to make a difference in these expansive sectors. And they were doing it through the forces of UNscale - where they could rent space and functionality in the cloud, a payments platform, access to highly qualified potential customers and more actionable data than ever before, giving them the same capabilities as large, scaled firms but without the bloat.
I wrote UNSCALED to capture how the forces of unscale are creating opportunity and innovation across every sector no matter how well established.
InfoQ: For whom is this book intended?
Taneja: UNSCALED was written for anyone who wants to learn more about contemporary forces in innovation - both the positive and the cautionary. The book is resonating well both with entrepreneurs and with anyone who is thinking about how data, AI and algorithms have the potential to impact society.
InfoQ: How do you define "unscaling"?
Taneja: There are really two parts to the unscaling phenomena. The first being an entrepreneur’s ability to rent instead of building the services - everything from office space to payment platforms. The second, which we’re just seeing emerge now, is the impact that the access to data and increasingly sophisticated AI is having on the pace and specificity of innovation. Early stage companies can learn more and iterate faster to create better products and solutions for smaller, more qualified markets.
InfoQ: Nowadays it's easy to rent office space and computing power. What other things are there that startups can rent, and how can that give them an advantage?
Taneja: SaaS solutions for everything. Payments platforms, cloud hosting and storage, document handling. And, everything HR-related from hiring to reviews, payroll and benefits can be done as a service. And they rent ad space that’s far more sophisticated in targeting qualified audiences than has ever been possible before.
InfoQ: What are recent developments that are causing Artificial Intelligence (AI) to break through?
Taneja: Easy: enormous datasets and computing power that has benefitted from decades of progress as defined by Moore’s Law.
InfoQ: How do companies apply AI to make products which better meet customer demands? Can you give examples?
Taneja: I’ll give an example that we describe in detail in the book: Livongo Health. There are 30 million diabetics in the U.S. alone, most of who manage their disease by tracking their blood sugar levels on a daily basis and checking in with their primary care physician a couple of times a year with a spreadsheet of numbers to look at. It’s a manual, backward-looking process for the individual that keeps them continually thinking about their disease.
Through the use of data and algorithms, Livongo takes some of that weight off the individual. Using algorithms against huge datasets, Livongo is able to recognize patterns, predict acute episodes, and help people better manage their disease on a daily basis. Through a connected glucometer, Livongo helps identify when an individual’s metrics are trending negatively. The app will suggest actions the person can take to avoid dangerously high or low blood sugar. And, if the situation is serious, a diabetes counselor will call the person directly, potentially helping them avoid a trip to the ER.
AI-powered apps will help people live healthier, more active lives and enjoy greater peace of mind.
InfoQ: What do you expect that AI will bring us in the future?
Taneja: That’s a very broad question. I think immediately, we’ll see AI help hone and improve products for groups of consumers — like the Livongo example. As more data is collected and algorithms become more sophisticated, the product feedback loop will shorten, iteration will happen faster and we’ll see some significant leaps forward in predictive health applications, autonomy, adaptive learning and many, many other areas.
I also think we’ll continue to grapple with the questions of the societal and ethical impacts of implementing AI across society-level services (content, commerce, communication, education, finance, government) for many years to come.
Finally, I believe that artificial intelligence will continue to augment human capabilities long into the foreseeable future. While advancements in automation will continue, we are a long, long way off from AI wholesale replacing people in the workforce.
InfoQ: What benefits does unscaling bring?
Taneja: As I mentioned earlier, forces of unscale make it possible for small teams of people to take on the incumbents in some of the most recalcitrant sectors. Being able to rent computing services in the cloud meant Stripe could build a global payments system without standing up their own server farm. It meant Gusto could use social platforms to find the exact right SMB audience for buying their HR solution. And it means companies like Color Genomics and Livongo can use AI to create products to create affordable, personalized, and preemptive and supportive healthcare that the scaled providers just cannot offer.
InfoQ: What about large companies, is there still a place for them in the future?
Taneja: Absolutely. There’s an entire chapter devoted to this in the book. To date, established corporations have used acquisitions to keep pace with the upstarts, which is a valid way to do business. But, they’ll need to adopt the Unscaled mindset if they want to compete directly.
About the Book Author
Hemant Taneja is a managing director at General Catalyst and is the author of UNSCALED, a 2018 book that explores topics of technology and society. Taneja joined General Catalyst in 2001 and launched its Silicon Valley office in 2011. His investment portfolio spans industries and includes companies such as ClassDojo, Color Genomics, Grammarly, Gusto, Fundbox, Livongo Health, Samsara, Snap, Stripe, and TuneIn.
