Co-authored by Sherman Lee and Rahul Vishwakarma
If we were to ask you to gather a group of three people to carry large stones block by block to build an Egyptian Pyramid, would you think we’re crazy? What if we gave you an infinite amount of time? And if you and the rest of your group are physically very strong? We’re going to guess you would outrightly call us stupid. It’s still a near impossible task.
Now let’s make a slight change to the instructions. Each person can call on two of their friends.
This small change is what made the impossible into a reality.
How AI came to be something that needs you
Pre AI Winter was all about gathering as much ammunition (read “IP in AI”) as possible and selling AI as Pepsi. But this parade hit the wall and everything stopped. This gestation period was necessary for the maturity of the technology and now it’s stronger than ever.
In the early 2000s, AI engineers had to have other skills to get a job. Eric Schmidt jokingly once said, “We didn’t even know that we had AI experts at Google during that time.” By the end of the last decade, efforts started to fall in place, and the innovation was growing exponential like never before.
Deep Learning became the common dinner table talk. Even people who didn’t understand what’s Deep in Deep Learning. Big companies have come to this realization that they need to open up all their research and technology for the masses.
Phase 1 — Advanced techniques in AI/ML (Deep Learning) got open-sourced and from AI researcher came into the hands of AI practitioners via numerous papers being published.
Phase 2 — From practitioners came into the hands of explorers and software developers via implementations on Github. This lead to the chain of events, from better well structured and optimized ML/DL frameworks like Torch, Theano, MXNet, Caffe, and of course TensorFlow.
This started the proliferation of incorporating AI techniques into real world scenarios and gave rise to even simpler to use Meta-Frameworks like Keras, Chainer, and so many more. And if this wasn’t enough, there are hundreds of AI-as-a-service companies distributing APIs to a very niche market. Incorporating machine learning in every way possible to game up against the rising competition is exponentially rising in demand.
Proliferation of AI
There are a plethora of use cases to apply artificial intelligence and machine learning to automate parts of your business. Now we’re not talking about completely taking away the jobs from humans. It’s about augmenting existing workflows to make things efficient and allowing humans to do their jobs better.
During the glory days of Yahoo!, there were tons of editors whose job it was to find related articles to any specific article you were reading on their network. Before the rise of automation this would be done by humans who would find articles based on specific categories, tags, etc. This was a lot of manual labor just like how websites used to be manually categorized into the Yahoo! directory.
So we implemented a basic unsupervised machine learning algorithm – specifically clustering. This grouped together all the articles that were closely related in topic so that you can follow news as a story from several sources. At least 5 research scientists and engineers worked on the problem back in 2010. Today, this could be done by one programmer leveraging a platform like Mateverse.
Customer service and support are also making great leaps in innovation. Companies like Amazon who have an insanely large amount of customer inquiries use bots to develop life-like replies. Only once the request becomes too difficult does a human take over. This means that companies can automate on average the first 5-10 replies to a customer. A customer support specialist can take over after retrieving all the required information. Thus, allowing them to serve more customers in a shorter amount of time.
Marketing teams are usually quick to want to adopt machine learning into their business. They have tons of data to work with and so many spreadsheets they’re using to manage everything. For example, who doesn’t have some sort of dashboard tracking every single post they make on Facebook with amount of likes, comments, and clicks correlated to the time and topics of the post?
Marketing sees a clear benefit to using their internal data to reach a large amount of people. From the type of Facebook ads that perform best, to potential blog posts that might go viral, to predicting which target audience might become your best customers. Getting started has never been easier. You can easily add a bot like Rocco to your Slack channel and leverage artificial intelligence to help you create social media posts that will get the most engagement.
Your sales team will love you if you implemented something to help them close more deals. Help calculate lead scores and find the deals that are likely to close soon. Help them uncover email templates that get the most positive replies. Help them craft responses so they don’t need to manage their inbox like crazy.
Taking some instances from the movie Her, the protagonist Theodore leveraged AI techniques to help compile the best of his letters into a book which a publisher accepted. The days are not too far from human skills being augmented a thousand times. We are and will be able to automate chores which require some rationale. Companies big and small today are trying to make that available as soon as possible. It’s primitive, but we are going strong.
I wish you the curiosity to be the explorer leading the way.
If you liked this article, we have great news for you! We will dive deep into when and how to apply data mining, statistical learning, machine learning, deep learning, and artificial intelligence.
Our goal is to cover every industry possible so even beginners in the mass market can start implementing this technology into their projects, their teams, and their organizations. We don’t believe you need 10 PHDs, 25 research scientists, or 15 data scientists on your team.
A highlight to what you can expect in coming days:
- How AI is amplifying Marketers’ skills, and tools they use.
- Sales on Steroids. ML techniques and tools some of the successful sales teams use.
- Even Content writers.
- How doctors and healthtech is leveraging AI.
- E-comm being the apple of eye for the most, has tonnes to use from. Which one’s are the best?
- Bots and conversational interface. How best you can build them for your business.
- Edge computing for the folks in Hardware.
- Above all, what are those right resources with which you can become a DL engineer yourself, without even spending a penny.
P.S.: This is not a marketing trick. We genuinely believe in building the community. What we are/will be sharing is something that we use and the knowledge we have gained in our years of experience.
We want you to follow along. Add yourself to the mailing list. We’d love to meet you and chat all things artificial intelligence and machine learning.