The idea in a nutshell: Bots are simple software applications which will automate tasks currently done by humans. They seem likely to succeed in taking jobs from people because they are the easiest solution to customer problems. Talking to a bot to get an answer is easier than opening the app and doing it yourself. This is the first step towards Artificial Intelligence rollouts and the effects seem material.
The motivation behind all this bot stuff ?
When Google change their primary strategy from Mobile First to AI first, I guess it’s time to start thinking about AI (Artificial Intelligence.) Because they’ve already finished.
As I’ve covered in a bunch of other articles in this blog, Facebook, Microsoft, Amazon and a collection of other huge tech companies have also started making significant investments in the area. Many of them are focussing on creating virtual assistants too. The revolutionary impact AI is going to have on the future of Digital, starting with bots, is beginning to become a bit clearer.
Bots (can) use artificial intelligence techniques to ‘learn’ how to reply to questions
Bots are ‘just’ ‘simple’ software programmes. Bots (can) use AI to assist in the way they answer customer queries. The point of bots, is to learn the way queries can usefully be responded to. They then do that automatically, all day. They’re much cheaper than people to employ. The idea is that bots do all the hard work, answering the simple, repetitive questions that most people ask. That then frees people up to do higher value work.
In the context of a phone company then, bots might work with people to help them recharge their prepaid phone plan so they can start talking again which makes up most of the calls in some call centres.
The people who were doing that work are then freed up to do higher value tasks. Again, in the phone company call centre example, people who were spending time recharging users might be freed up to proactively manage a customer account or convincing people not to leave the company if they were pissed off.
The key word is learned
Bots and AI are a big deal these days as a result of the interaction of a number of different technology factors. Cheap computing and storage available through the cloud, huge, also cheap data storage solutions and improvements in Neural networks to learn from the patterns they see. In some ways, they work like the human brain does.
Bots, Virtual Assistants and AI are all very similar
Virtual assistants (in this context) are just advanced bots which have been taught to deal with a broader array of tasks.
Bots are useful because they use the simplest user interface there is
The value of bots comes from the ease of use the provide. Bots are easy to use for two reasons. First of all, talking is kind of the simplest user interface there is. Right from day one, it’s how parents deal with their kids. It’s how we relate to each other, mostly, in shops, bank branches and schools. It’s even how they imagined the future on Star Trek.
Secondly, bots are in the places you go. If you’re in Facebook, bots will soon answer your queries. If you’re searching in Google, same deal. Google have been quite up front about their view that the technology is fading in to the background. What’s important is associating the user intent with the outcome they want on the internet. The interface to achieve that can be anything you like. And people tend to use the easiest, laziest interface they can find.
Why are Microsoft and Facebook so Interested in Virtual Assistants
One school of thought on why Facebook and Microsoft would come up with such similar strategies (focussed around bots and AI) is that they are trying to upset the apple cart and change the duopoly that Apple and Google have going at the moment in technology on who makes money out of the internet.
I guess it’s obvious that all the money goes to the creators of the platforms these days. Microsoft (using Skype as their platform) and Facebook (using messenger) want to make it as easy as possible for people to build their bots on their systems.
At a technical level, it seems likely they want to own the platform so they can aggregate the content of the discussions and improve their own Virtual Assistant / agents. They’ll have so much more data, so many more conversations and such a plethora of experience, being the platform for bots, that their bots will be streets ahead. That’s why it seems likely that the virtual assistants, the most advanced and intelligent bots, will come from them.
If Microsoft and / or Facebook do figure this out and get traction behind their bots (and some estimates suggest that there are already as many as 10,000 coding houses designing bots for Facebook already) they could change the fundamental interface we have to the internet. Instead of typing in to a search box, we could be talking in to a (Microsoft or Facebook) bot’s ear.
There are some interesting effects which might accompany this sort of change but, at a headline level, ownership of that interface (and finding a way to monetise it) would certainly be an improvement for either Microsoft or Facebook’s position relative to Google.
When bots & voice interfaces won’t do
It’s interesting to note that even on Star Trek, there were interactions with technology which did not involve speech. On the main control deck where they looked out the window at space and the captain gave all the decisions, there were a lot of people trying to work at one time. They used touch interfaces on desktop consoles to access information and inform the computer of decisions: like going to warp, putting something on main viewer or checking the efficiency of the core (yes, I watched it a lot.)
Similarly, it seems like there are going to have to be touch screen interfaces even in a world revolutionised by bots. If I ask my virtual assistant what my bank balance is or how much I am going to get paid this month, I might not want him / her to say it aloud. In an environment with other people, like a lift, car or office, I might want to message either my Virtual Assistant or wife silently rather than verbally.
Summing up the bot life
It seems likely that bots (as a first step towards AI solutions) are going to have a material impact on our lives and businesses in the near future.
Beyond the immediate and more tangible effects, I wonder what effect the availability of an infinitely patient ‘friend’ in the form of a Virtual Assistant might be for those who currently feel lonely. The effects of loneliness are terrible and a growing part of life. It’s literally true to say that people die earlier when they’re lonely. If you have a best friend bot, perhaps you will live longer.
Not that all the effects will be positive. I work in a building where, when you try and get through security, you present your security pass to a machine. The machine then tells you which lift you should get in to. Every day I notice that people follow the same procedure (touch their security card to the sensor, get told what lift to walk to) and then immediately forget the lift they were sent to. They have to go through the process again manually. This is slightly more effort than if they had just used the old process which was to hit the button for the lift. If bots make life easier for us will they also make us more stupid ? Is this progress ?