Who will own the future of intelligent chat bots?

On one hand, there are internet giants such as Facebook, Google, Amazon, Alibaba and Tencent. These companies have two distinct advantages:

  • Cash that can be invested for the long term, unlike venture capital, which must aim for an exit time. That means they can really invest in something for which the short term return is highly unlikely;
  • Data – enormous amount of data that allows them to train their machines much faster and more effectively;

On the other hand, there are VC-backed maverick companies. These companies have only one aim for their existence: to create AI in the same ranks as “Her” or “The Terminator” (more of the former).


Compared to the giants, these companies do not have as much data to work on – and they are trying hard to get around it.


Some work relentlessly to acquire partners who have that data; some scrape whatever they can find in public (and sometime not so public) domains; while some others hire an army of editors to fabricate simulated data.

These companies also work hard on the funding front. They are aggressive in PR as well – creating public and investor awareness, and thus more support in terms of funding.

Sprint is over, marathon now

They know that to reach there, they need years of research and trial/error. And what is important: to survive these years.

In order to survive, you need to be top of the brass, giving investors confidence that if and when this happens, you are there to reap the most lucrative rewards.

Many investors know that the real intelligent human-machine interaction, if cracked, could bring MASSIVE value. Many are also aware that the short term P&L could be quite dismal. They have to back the top player(s).

That’s why in China there used to be more than a dozen companies competing for the prize. This year, barely three among them are still operating.

Among these is Emotibot, which was started by a Microsoft veteran in the space. The company aims to detect, and effectively interact with, human emotions.

Emotibot spent the past 18 months developing the base technology, covering text, facial and voice. Now comes the real test of actually deploying this into e-commerce, finance and other industries.

Thus far investors have put confidence to into them, but whether this confidence will persist really depends on how fast they can find good, profitable use cases.

Oh yes, and they have to make sure that the Chinese language chat bot is politically astute, and not picking up some of the ‘bad’ comments online during the learning process.

Sounds easy for a human being, but incredibly difficult for a machine, especially with unsupervised learning.

No room for small fish

In places like Beijing and Singapore, we also see a big school of small ‘AI’ companies, offering niche services and often with fewer than a handful researchers.

Let’s face it – these small fish will mostly perish under the pressure of big sharks.

Yes, some of them might get acquired – but what is often acquired is the team, not the technology, therefore, without much premium.

Occasionally there will be ones cracking some really smart algorithm – alas, algorithms are not as valuable as we might think.

The real differentiator is NOT algorithm

Yes, contrary to many of us might think, algorithms are not the most difficult part. All the big advancements in AI are pretty transparent and as long as you have enough good scientists (which are in fact in short supply), you can figure out algorithms one way or another.

One thing though is equal (and crucial) for both internet giants and startups: tagging of data.

Similar to big data industry, AI requires clean data for the machines to map, and learn.

The human touch

No matter how fancy the technology is, the customers you serve are human beings. Therefore, companies should always remember what make a great customer experience.

A few months ago, I was enquiring about an ecommerce order through Livechat. Everything was going very smoothly and a bit too fast, so I asked “Are you a human, or a robot?” The answer was “Lol. I’m a human. My name is Melvin”. That made my day in the world of technology – even if Melvin were in fact Alexa or Siri.

A dose of good old human touch never fails to leave a great customer experience.




Thanks for reading The Low Down (TLD), the blog by the team at Momentum Works. Got a different perspective or have a burning opinion to share? Let us know at [email protected].


Previous articleChatbots and AI- Part 1: not married yet, but soon
Next articleRein in the dynamics between disruption and regulation
Jianggan Li is the Founder & CEO of Momentum Works. Prior to founding Momentum Works, he co-founded Easy Taxi in Asia, and served as Managing Director of Foodpanda. The two years running Rocket Internet companies has given him a lifetime experience on supersonic implementation, and good camaraderie with entrepreneurs across the developing world. He holds a MBA from INSEAD (GMAT 770) and a degree in Computer Engineering from Nanyang Technological University. Unfortunately he never wrote a single line of code professionally - but in his first job he was in media, travelling extensively across Asia & Europe, speaking with Ministers & (occasionally) Prime Ministers. Apart from English and his native Mandarin, he is also fluent in French and conversational in Cantonese & Spanish. He tried to learn Latin (for three years) and Sanskrit (for six months) as well. In his (scarce) free time, he reads, travels, hikes and dives. Pyongyang, Tehran & Chisinau are among the interesting cities he has been to.