Last week our CTO postulated how machine learning would disrupt transactions to have a ripple effect across industries. Today I posit that there are five technological changes that expedite the path to entrepreneurship.
Lowered barriers to entry
In the 1990s, it cost at least US$ 5 million to launch a product and reach 30 million users. Building anything remotely close to Careem in the 90s meant having to buy and manage your own servers, needing an entire floor of scientists to write your own machine learning libraries while A/B testing which library is the right for you, and churning out the code for flat HTML pages. You would then hire a DBA (database administrator), pay ten large for a slice of Oracle’s database and have to learn how it’s configured.
Today you can launch a product and potentially reach 1 billion users in under US$ 1 million. Instead of buying and managing your own servers, you would just have to use Amazon Web Services. Instead of hiring scientists for machine learning libraries, you would use the Google Prediction API. A/B testing would be handled by Optimizely. DBA’s would be replaced by Amazon RDS. What took decades in the past can now be accomplished in months. Soon it will take weeks. For proof look at the list of products being launched daily on
What took decades in the past can now be accomplished in months. Soon it will take weeks. For proof look at the list of products being launched daily on Product Hunt. We launched Tagxit, a machine learning camera app, in just a few months.
This technological change is the foundation of the remaining four below because time is money and speed to market is a critical factor for dominating customer mindshare early on. The aggregate value of building industry specific software for a range of verticals eclipses anything that Careem may support. In a world where execution trumps ideas, maturing off the shelf infrastructure will open doors for idea implementation by just about anyone. While software development is a barrier to entry today for solving problems, it won’t be in the near future, and it will open more doors for intrapreneurship than imagined.
As a result of the intrapreneurship mentioned above, the escalation of B2B vendors has been unprecedented for verticals and functions. The latter has ranged across marketing, sales, engineering, finance, recruitment, and operations. The abundance and choice for buyers has resulted in data silos and confusion. The technological change represents an opportunity for middleware companies like CX Cloud that act as a translator or a matchmaker of sorts because software’s are made with varying languages. Middleware helps tie vendors together for streamlined integrations and reporting. I expect to see no less than a dozen billion dollar middleware companies by 2020.
The internet for all
A year after founding Google X, Project Loon was formed. Companies that depend on internet users are going out of their way to be providers of internet, democratising the access for all while commercialising it. Remember Facebook’s attempt with Internet.org backfiring due to its limitations? By the year 2019, a company called OneWeb aims to launch a more than 600 tiny satellites designed to beam high-speed Internet down to Earth. Elon Musk’s Space X put 4,000 small, low-cost, disposable satellites into orbit last year. When billions of people suddenly gain access to the internet for the first time, the audiences can spring forth will be very different than what we face today.
As more software startups go to market faster than before and integrations occur, people will rely on machine learning to take out manual tasks from everyday life:
- Open sources frameworks and algorithms will offer startups the ability to syntactically analyse, filter, and generate data without assistance from humans
- Using Google Voice for dictation into text and vice versa
- Conversation bots to assist people with frequently asked questions
- Aipoly uses AI to give the blind sight with photo and video recognition.
When Amazon created Amazon Web Services, it was to turn a cost centre into an enormous revenue stream. I expect to see the rise of Labs as a Service, allowing the renting of idle equipment during downtime between projects. Leave the benchwork to the robots and let the humans design the experiments. Startups like FarmLogs are helping farmers run profitable farms, adding a data-driven process into agriculture.
The technological changes listed above will continue to mature, well past the year 2020, and will offer humans the leverage required to inherit feedback loops.
The views expressed here are of the author’s, and e27 may not necessarily subscribe to them. e27 invites members from Asia’s tech industry and startup community to share their honest opinions and expert knowledge with our readers. If you are interested in sharing your point of view, submit your post here.
Featured Image Copyright: niserin / 123RF Stock Photo