Which Ai Operating System (AiOS) to Choose for your Business in 2018
I often refer to the present state of the artificial intelligence as a modern Gold Rush 🤠 of opportunities for all of the Ai start-ups out there. You don t have to look for to see the how much artificial intelligence interest the most and the biggest tech giants are going all-in with Ai.
The artificial intelligence world right now is like a volcano erupting from all sides with new features, new market to service, Ai applications swirling all over the place. This high activity climate will cool down in a year or two, but until then, it will be hard to predict who will be the winners and the losers of this breath taking race.
Google and Facebook Slicing the Ai pie
For me, the future of AI has two components: the smartest AI will replace search engines and Google with Dialogflow and its President saying that Google is now an AI first company is a good sign of them leading this game. The other front runner to finish second and third is Amazon and Microsoft with their respective voice recognition Ai platforms Alexa and Cortana.
On the other side Facebook will focus on AI applications that are aiming to replace Android Apps and Ios apps, this is the Manychat, Chatfuel, Facebook and other text based platforms with limited Artificial intelligence capacities. In December 2017, a clear shift was into operation and it seemed like Google and Facebook decided how they will split their Ai pie:
Google will, occupy the Conversation as a Service (CaaS) and extensive Natural Language Processing (NLP) market with advanced machine learning.
Facebook will focus on the Android and iOS 2.0 like market with simple and intuitive chatbots that will access through your mobile phone applications using web views.
I am not saying that either are better than the other one, they are just different products aimed at different markets.
So what your artificial intelligence operating systems (AiOS) should you invest in? I created a list of 5 questions to help you figure it out.
Where are Your Customers — Nowadays most customers contact businesses online (that no surprise for most people). This is of course, but we need to dig a bit further: Do you sell goods on Amazon? Do most of your customers’ questions come from your Facebook Page? Does your business could is a good fit for Internet of Things (IoT) home devices such Amazon Alexa or Google Home?
A restaurant could benefit from a Facebook Messenger chatbot that handle the clients orders, delivery requests with PayPal or Stripe payment integrations, but a voice recognition NLP agent would not be appropriate in the same context because the Ai would not handle well five people talking to the agent simultaneously at the same table.
On the opposite side, I can imagine a world in which you will interact with your car with voice recognition through Internet of Things integrations and ask:
“Ok Google, suggest me the best Italian restaurant in Manhattan near of Chelsea Market.”
Which will be much safer then texting and driving by the way. 😉
Knowing where your customers are, and how they interact with your business is probably the most important question you need to ask yourself when selecting the appropriate AiOS for your business.
What Type of Ai you are looking for — This the most important, yet obvious question you need to ask yourself. Bots are new shiny toys and most people looking for a bot right now are like this:
#iwantitnow
But before getting too excited by bots, make a rational brainstorm with your people from your team both with marketing related expertise and business management experience. At this point you should not worry about any IT guys; focus on developing 2 o 3 business user cases of what is the main goal you want to achieve with your Ai solution. Here are 3 examples of this type of exercise:
- We want our Ai agent to be able to improve our customer support rating on our Facebook Page by being able to answer our 20 most frequent asked questions;
- We want our Ai solution to be a Facebook Messenger sales funnel that will reduce our CPC by at least 5%;
- We want a delivery Ai app that will increase our online delivery request by 10% in the next 3 months.
Do they have grey hair or not — My mom finally figured out how Facebook and Messenger work out with her mobile phone (sorry mom). I am sure she will eventually figure out how bots works, but I am sure she’s ready to go into a restaurant or in a shopping mall and interact with a voice recognition advanced machine learning app. If your customers are older or you are a financial institution, a real estate agency, you might be better to stick with Facebook Messenger bots, but if your customers are young and trendy, then you should considerate both AiOS.
Do you retarget your customers — I am a Google guy I admit it and a true believer of Natural Language Processing future (i.e for various NLP AiOS not just Dialogflow BTW), but if one of your business Key Performance Indicators (KPI)is retargeting customers that abandoned cart items or if you are a restaurant that want to reach a customer to return to your restaurant next week at lunch time, then look nowhere else than Facebook Messenger bots with either ManyChat or my with Chatfuel.
Natural Language AiOS are great but their strength is to recognize users basic to complex intents but they are not as good with retargeting customers based on their actions with advanced analytic tools.
Do you want Ai to replace tedious tasks — This is my favourite and I will always tell you when my heart overcomes my rational mind. If your Ai vision for your business is to be able eventually to replace a portion of your employees tedious tasks to make them focus on value added portion of their jobs unlike:
- Filling forms 🤢
- Responding to repetitive email queries 😡
- Finding a meeting room for tomorrows meetings 😧
- Answering repetitive FAQ questions that are on your web site, but still some clients don’t take ONE MINUTE AND A HALF to go through your damn web site FAQ section (anyone can relate to this one?) 😒
This is what NLP agents are best for! Whether you need a virtual assistant, a task reminder agent, a customer service relationship agent, a movie recommendation Ai solution, Conversation as a System Ai solutions are what you are looking for. Coupled with machine learning your CaaS agent will not only make some of the most tedious tasks, it will also learn and from its mistakes and become better with t ime. The more he or she will interact with your customers, the better he or she will become (notice the use of he and she, for me robots and wayyyy too cool to be called it. Beside that It movie was really awesome).
This particular user case might not interest a portion of the small and medium size businesses, but trust me, bigger corporations will be really interested in this user case in the near future, not only to reduce their costs, but also to allow their employees to have a better satisfaction of their work position which will increase their overall happiness at work and decrease their level of job related stress.
The Ai arm race just began, I expect a lot of new user cases and new features to be developed throughout 2018 and this is why I felt in love with Ai even though we have a “it’s complicated” 🙄 Facebook relationship status from time to time.
If you like this article please give me a couple of claps 👏👏👏👏👏 below people would be surprise how much a simple form of appreciation do make us Medium writers to have a great day just because you swipe your finger on your mobile phone and give us some gratitude. You can also share this article on social media for your Ai geek friends. 🤓
If you have any questions 🤔 I’ll be more than happy to answer the or if you have any Artificial Intelligence, Machine Learning, Data Science topic to suggest me, please write me an email 📧📧📧 to carldombrowski@hotmail.com.
you can also show your support Carl Dombrowski on Patreon by becoming one of my patron for my soon to be launched natural language proicessing video tutorials : https://www.patreon.com/bePatron?c=1461648