The Rise of the Bots

Much has been written lately about the rise of chat / social bots. One of the best views is Chris Messina’s widely shared post on how 2016 will be the year of conversational commerce. This article is not an attempt to summarize many of the key opinion makers in the space, it is focused on highlighting three counterintuitive trends around chat bots.

1) Humans in the Loop

Bots are all about machines, right? Not quite. While we have managed to create bots that pass the Turing Test they are more specific than general. Try conversing with a bot about anything and it is likely to eventually trip. Which is why we need data to continue training algorithms so they become increasingly more sophisticated. Lots of data. Consider that Facebook M, arguably the most visible effort on marrying messaging and AI, has humans in the background. One could argue that an approach relying on humans in the loop is expensive and doesn’t scale. Facebook’s bet though is you will get better results this way and give the company enough data to train M around user intent and actions ie eventually needing less human curation.

2) People will not want to talk to a machine

People almost universally hate automated systems when calling customer support. The thinking is even if we improve the systems dramatically they will still be subpar humans and humans will prefer talking to other humans. Nothing could be further from truth. Slack is overrun with bots; TechCrunch just joined the fray and created its own bot on Telegram; even Disney has joined in by incubating a startup around a Miss Piggy bot… behold the future of brand engagement.

Yes, we will need to cross the chasm between early adopters and the mass market. But it’s no different than the adoption of practically every (communication) technology before it. Consider how people initially resisted the idea of buying a smartphone partly because they felt uncomfortable being reached anywhere — eventually we developed features and an etiquette on how to communicate in an on-demand world. It will be similar as we transition into a world where we converse directly with machines.

3) It’s just like in Asia

Nope, this is different. WeChat (China), Kakao (S Korea) and Line (Japan) are indeed much ahead of the dominant platforms in the West when it comes to an integrated suite of features, whether it be paying through your bank account or hailing a cab. The core work around bots though is coming from Silicon Valley and Israel and the paradigm here is not on integration but on intelligence. It’s not about creating an icon or link within the messaging platform to another service, it’s about having an AI that you can converse, to learn, debate and execute your actions. There is value in both integration and intelligence and they may indeed converge when the dominant interface evolves from text to voice. That day, like in the movie “Her”, may not be too far indeed — but that’s subject for a future post.

Have an opinion around bots? Fire away in the comments.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.