ProjectBlog: How AI is changing innovation

Blog: How AI is changing innovation

Some time ago I read an interview with a Spanish serial entrepreneur. While most of his answers were based on the usual start-up jargon, one specific line caught my whole attention, something in the line of:

“Sometimes it’s not better to be the first, it’s better to arrive last but close the door to the competition when you arrive”

We always look at Silicon Valley trying to find the latest innovations, new business models based on fancy technology but, is this the right way to look for innovation in our corporate world?

Geoffrey Moore talks about three returns on innovation: differentiation, neutralization and optimization. According to Moore, companies should focus on:

  • “Unmatchable” differentiation.
  • “Good enough” neutralization.
  • “Save to invest” optimization.

If you think about those categories, Silicon Valley and its start-up ecosystems comes rapidly to mind when thinking about unmatchable differentiation. They create new technologies and new business models from them. Is this going to change with AI? According to Kai-Fu Lee and others, yes.

Chinese technology companies have been always considered copycats, producing cheaper versions (optimization) of technology created in America (differentiation). But what if China had an “unmatchable” differentiation in AI? In fact, they do, and it is not based on technology. It is based on data. Huge amounts of data. Just take the example of Didi vs Uber:

Comparison of Didi Chuxing and Uber World. Source: data from Roland Berger, Recode, China Daily, Crunchbase.

During 2018, and only in China, Didi doubled the total number of rides made by Uber in the whole world. Also, the have 7 times more drivers. That means data. About customers, about employees, about how cities behave.

Still, one could argue that it is good to have more data, but it is even better to have the best researchers and Data Scientists around. That should be the key to stay in the edge of differentiation, right? Well, according to AI experts, not the case. It turns out that well known models trained with more data perform better than the most advanced ones with less data.

This week I found an article discussing the new Huawei’s technology campus in China. The article focused on how they recreated 12 European “towns,” with names such as Paris, Verona, and Bruges. Research buildings are also modeled after famous castles, palaces, and more. As I stated in this tweet, I think we are missing the point if we focus only on how they have copied European architecture.

Although Huawei’s example might seem irrelevant, I think it can show us a lot on the Chinese bet on leading Artificial Intelligence. They are attracting and generating really good talent, and all of the AI ecosystem is being supported by a very ambitiuous plan by their Goverment.

We can have a very good example of this combination of efforts in the city of Taijin, which has invested 16$ billion in creating its AI hub. The Chinese government is clearly betting on AI, and they have the money to create not only technology hubs, but also new cities from scratch that can be home to millions of people, therefore creating even more new valuable data.

According to Andrew NG, they are differentially prepared to fulfill the virtuous cycle of AI.

Are we ready to compete in this new era of innovation? According to Kai-Fu Lee, we are not.

Should we all learn Chinese? Hopefully we don’t need to, as long as AI will surely serve as a translator.

Source: Artificial Intelligence on Medium

Leave a Reply

Your email address will not be published. Required fields are marked *

Back To Top

Display your work in a bold & confident manner. Sometimes it’s easy for your creativity to stand out from the crowd.