Because the CEO of an organization the place machine and studying and AI play an necessary position, I’m at all times interested by seeing how different organizations select to implement these applied sciences into their group. It’s fascinating to see how Uber vs Lyft has been battling for supremacy — via tellingly completely different approaches to AI.
How each Uber and Lyft have moved ahead says rather a lot about their total targets as an organization. Because of their current IPOs and the quantity of data that has been publicized in consequence.
On the face of it, Uber and Lyft look comparable.
Uber and Lyft seem like pretty comparable firms: each have made their names within the ride-sharing enterprise whereas producing billions of in income. Take a better look, nonetheless, and it turns into obvious that every firm has its personal very completely different method to success. They every have their very own technique of making use of AI as a way to obtain it.
Uber, because the bigger firm of the 2, has extra sources at its disposal, and due to this fact has the flexibility to spend extra time constructing out a complete machine studying platform and platform.
Lyft, alternatively, could be smaller however the firm nonetheless has entry to an amazing quantity of buyer and driver knowledge that it makes use of to optimize its personal ride-sharing platform.
Regardless of their dimension variations, they’ve additionally publicly defined the variations of their company philosophies, and people variations have carried over into their approaches to AI.
A lot of the knowledge accessible concerning the two firms.
The dear data come from their respective SEC filings, made previous to IPO. These filings are a treasure trove of data for anybody trying to study extra about how firms resembling Uber and Lyft construction their companies. The place do they assume their biggest worth lies?
As ZDNet author Larry Dignan notes, Uber’s filings “[make] it clear that its knowledge science and algorithms are the important thing to its market applied sciences,” whereas Lyft’s IPO paperwork don’t really point out synthetic intelligence instantly.
The Lyft angle.
What Lyft takes care to focus on, nonetheless, is the information it has collected “from over one billion rides and over ten billion miles pushed,” which has been used to develop machine studying algorithms and inform knowledge science engines.
In accordance with Lyft, the insights from all this data are used “to enhance the product expertise for riders. Riders are offered with customized transportation choices,” in addition to to “anticipate market-specific demand” and “create custom-made incentives for drivers in native markets.”
Lyft focus is much less on innovation and extra on the best way to good the service that it gives.
Primarily based on all of this data, it’s clear that Lyft sees machine studying as a method to enhance their present providers. Uber – as I’ll discover shortly – has a special take.
What does Lyft take into account to be the core parts central to the corporate’s success? Overwhelming focus is on its management, core values, and the relationships it has with staff, drivers, clients, and companions.
Uber, alternatively, emphasizes its “deep expertise benefit.”
Uber appears on the quite a few proprietary methods it has constructed, and its acknowledged purpose for the approaching years is to make use of this expertise “to redefine the huge meal supply and logistics industries.”
For Uber, expertise will not be merely the means for offering a greater ride-share expertise – it’s additionally the chance to develop into different industries and utterly rework them.
Uber leverages synthetic intelligence and machine studying.
Uber makes use of machine studying to foretell demand for rides, in addition to decide optimum routes based mostly on the time of day, site visitors and climate circumstances, and every other elements that would have an effect on motion.
The corporate has additionally built-in AI into just about each side of the enterprise, from customer support to fraud detection to advertising spend to the onboarding course of that drivers must undergo.
Uber has its personal AI analysis crew, whose findings are commonly revealed and offered at conferences all over the world, and has launched open-source software program for outsiders to make use of.
In accordance with Uber’s SEC filings, the corporate makes use of machine studying in many various methods. Using pure language processing has, within the firm’s phrases, helped to “simplify and improve interactions” on their platform. It permits the corporate to economize by forgoing human customer support brokers for automated ones.
Uber makes use of pc imaginative and prescient to confirm drivers’ licenses and different necessary paperwork (once more lowering the necessity for human interplay), and what it phrases “sensor processing algorithms.” These assist to enhance its location accuracy in crowded areas.
Wanting past the rideshare enterprise, its algorithms are used to estimate meals preparation and arrival occasions, in addition to to generate customized suggestions based mostly on an individual’s ordering historical past.
Uber does rather a lot with machine studying.
This isn’t to say that Lyft is detached to the expertise; relatively, it speaks to the very completely different long-term methods that every firm has determined to make use of. Uber, one might argue, has positioned itself as a tech firm, whereas Lyft’s mission, to “enhance folks’s lives with the world’s greatest transportation,” is extra targeted on the non-public impression that the corporate could make.
It’s for that reason that Lyft claims to make use of machine studying to make the expertise for shoppers and drivers higher, whereas Uber touts the elevated effectivity and automation that utilizing machine studying can convey.
Which method is best?
That depends upon your definition of success. On the one hand, Uber is a a lot bigger firm than Lyft, with a present market cap of over $68 billion (in comparison with Lyft’s $15.5 billion) and operations in over 70 international locations.
Then again, that doesn’t appear to have had a optimistic impression on its inventory value, which has fallen considerably because the firm went public. It additionally doesn’t assist that Uber has persistently made headlines for all of the improper causes, from endemic sexual harassment to a gaffe-prone CEO.
The query then turns into, is it extra necessary for a corporation to have a “human contact,” or to concentrate on expertise first and folks second? The reply will in the end be determined not by the monetary markets, however by customers themselves.
Jeremy Fain is the CEO and co-founder of Cogntiv. With over 20 years of interactive expertise throughout company, writer, and advert tech administration, Jeremy led North American Accounts for Rubicon Undertaking earlier than founding Cognitiv. At Rubicon Undertaking, Jeremy was answerable for international market success of over 400 media firms and 500 demand companions via Actual-Time-Bidding, new product improvement, and different income methods, guaranteeing interactive patrons and sellers might take full benefit of automated transactions. Previous to Rubicon Undertaking, Jeremy served as Director of Community Options for CBS Interactive. With oversight of a $30 million+ P&L, Jeremy was answerable for improvement, execution and administration of data-driven options throughout CBS Interactive’s community of branded websites, together with viewers focusing on, personal trade, and customized viewers options. Previous to CBS, Jeremy served as Vice President of Trade Providers for the IAB, the place he formed interactive trade coverage, requirements, and greatest practices, resembling the primary VAST commonplace and the Tc&Cs three.zero, by working each day with all the most important media firms in addition to all of the company holding firms.